How does institutional context shape work-related functionings for regular and self-employed workers? A contextualised application of the capability approach to Belgium, France and the Netherlands

Abstract

Purpose

According to Sen's theoretical framework of capability (1985), individuals reach their full potential once they have the freedom, intended as the set of functionings at their disposal, to do so. However, many critiques have been developed against the lack of embeddedness of the capability approach in social and political relations and structures. In this article, the authors investigate the influence of three institutional contexts (Belgium, the Netherlands and France) on the respective work-related functionings of self-employed and regular workers, with a focus on human capital investment and institutional support offered to them.

Design/methodology/approach

Data from the European Working Conditions Survey (EWCS) are used to highlight similarities and differences in building work-related functionings for regular and self-employed workers. A regression analysis is provided at the country level.

Findings

In the three labour markets, the authors find that the building of work-related functionings is more successful for regular employees, especially as regards institutional support. Self-employed workers, on the other hand, need to rely on their individual capability as regards employment protection and human capital investment. However, the authors find interesting differences between the three institutional contexts. In both Belgium and France, self-employed workers are subject to higher instability in terms of changes in salary and hours worked, whereas atypical work is better positioned in the Dutch labour market. The Netherlands is also characterised by a less significant gap between regular and self-employed workers with respect to participation in training.

Originality/value

In this article, the authors contextualise Sen's (1985) theoretical framework by taking into account the institutional differences of labour markets. In particular, the authors provide a novel application of his capability approach to regular and self-employed workers in an economically relevant European area.

Citation

Focacci, C.N. and Pichault, F. (2023), "How does institutional context shape work-related functionings for regular and self-employed workers? A contextualised application of the capability approach to Belgium, France and the Netherlands", International Journal of Sociology and Social Policy, Vol. 43 No. 13/14, pp. 36-61. https://doi.org/10.1108/IJSSP-12-2022-0312

1. Introduction

The current labour markets are characterised by precarity (Greer, 2016), often due to insecure flexibility (Berton and Richiardi, 2009); digitalisation, due to the current information and communication technology (ICT) revolution (Perez, 2016); and lack of proper contractual regulation, due to the accelerated rise of the gig economy (Pangrazio et al., 2021; Woodcock and Graham, 2020; Wood et al., 2019). In parallel, occupational inactivity has increased everywhere in the world in the recent decade, causing individuals to take on jobs that do not match their skills, or alternate between atypical job contracts and unemployment (Auray and Lepage-Saucier, 2021). The 2008 crisis, for instance, caused job deterioration, distress in the workplace and general lower levels of well-being in workers (Ogbonnaya et al., 2019).

The rising number of NEET (not in employment, not in education and not in training) individuals in the European Union territory, for instance, has led states such as Italy, Spain and other Southern areas to create institutions able to respond to the occupational insecurities that a passive welfare system granted to individuals (Crépon and van den Berg, 2016). In her article, Focacci (2020) recently showed how on-the-job training programmes can help individuals otherwise isolated from the labour market to achieve occupational stability in addition to employment. These potentially allow for better matching of skills and subsequent quality of jobs (Brown and Koettl, 2015). As a consequence, policy makers are, now, abundantly resorting to active labour market policies; namely, programmes designed at the national or international level and aimed at the reskilling and reinstatement of the extant or potential labour force into the working segment of the population. As passive labour market policies such as unemployment benefits and other forms of financial assistance fail to prepare individuals to the globalised labour markets, new attention has shifted towards policies of employment that require human capital investment. The latter can also help overcome involuntary pro-cyclical self-employment, with workers insufficiently supported by ad hoc training measures; as well as hybrid forms of employment (Murgia et al., 2021).

At the same time, more and more studies highlight the impact of aspects of individuals' lives not related to work on their occupational status. This is in line with the argument that work-life balance plays a crucial role in overall life satisfaction (Sirgy and Lee, 2018) and that a separation between work and non-work role segmentation preferences (Methot and LePine, 2016) suffers from an actual effective implementation. Certain categories of professionals, for instance, can be easily influenced by their peers when making a decision with respect to where to work, what type of job contract to sign and whether or not to participate in a training programme that could lead to employment, or a change of job (De Clercq et al., 2021; Focacci, 2020). The same applies to other segments of the working population based on their gender, ethnicity, education, sector of employment and role in the company. Gender inequality in the labour market, for instance, predicts women taking care of members of their family more often than men (Lott and Chung, 2016), therefore disfavouring their occupational activity. Folbre (2006) wrote on the possibility to define care as a way to enhance and better understand the impact of economic development on women. With respect to employment sector, the category of university employees usually fails in setting defined boundaries between nonwork and work, boosting perceived levels of work-life conflict (Buchanan and Boswell, 2006).

It is, thus, evident that there exists a connection between what the individual is able or unable to do in the world of work and his social and private life, outside of work. With respect to this, Amartya Sen's capability approach (1985) considers capabilities as different combinations of functionings that can be achieved by the individual. In other words, these determine the individual's opportunity to generate valuable outcomes in consequence to the freedom the individual possesses in order to achieve said outcomes. Sen's approach is, therefore, fundamental to shed light on what makes the realisation of an individual, intended as the entirety of what a person is “capable” of doing or being, possible. While income and wealth more in general, contribute to measuring the individual's “economic” wellbeing, Sen's moral framework suggests that what truly matters is the individual's freedom to live her life according to what she actually values. To do that, the individual needs to be provided with the right functionings; and these may be also provided in the world of work. This is in line with the argument that workers' well-being is of relevance for work and, as a consequence, human life (Budd and Spencer, 2015). In a recent study, Bueno (2022) wrote about the necessity to make work a capability-enhancing activity, independent of whether it is productive or not.

Given the social, economic and moral power that working conditions have on the development of the individual as a person, we think it is crucial for policy-makers to understand how certain types of characteristics that we define within a context of work-related functionings, i.e. characteristics that are needed by the individual to express her full potential at work, are shaped by institutional context. Many critiques have been formulated against the too individualistic perspective in which the capability approach has been developed (Robeyns, 2005; Godfrey-Wood and Mamani-Vargas, 2017), with few considerations of “the embeddedness of economic action in social and political relations and structures” (Leβmann, 2022). Most suggestions to overcome this problem relate to the concept of collective capabilities, including the institutional level (Ibrahim, 2017). We follow this perspective by exploring the extent to which various institutional contexts can provide workers with different capabilities. In this regard, we contribute to the literature by providing one of the first application of Sen's capability theoretical framework to the working conditions debate, contextualised to the institutional differences present in the countries under investigation. Our main hypothesis is that different institutional characteristics —including the country's union rate, density, coverage rate, collective bargaining, social rights of self-employed workers, implemented active labour market policies— are likely to affect the gap in capability formation between self-employed and regular workers. This, on its turn, will have some influence on the experienced quality of work.

In other words, applying Sen's capabilities to self-employed workers in different institutional contexts helps reintroducing structural effects in a theory often considered as based on individual agency (Leβmann, 2022). The focus on self-employment is justified by the growing number of highly skilled workers in this category, the degree of autonomy that characterises their employment, and the new challenges this poses for the modern labour markets. According to the recent analysis by Pichault and McKeown (2019), the work of independent professionals has spread so extensively in both numbers and sectors of reference that it is no longer possible to talk about atypical employment. In fact, self-employed workers may enjoy the benefits that come from flexible definitions of their work status —ranging from independent contractors to almost regular employees— and work content —distinguishing between workers with high levels of discretion in workload and work pace and workers with standardised work processes. However, whether the autonomy they experience in terms of working conditions is beneficial to their social and economic wellbeing and fruitful to their work and output highly depends on how such working conditions are supported. On this subject, Gries and Naudé (2011) show that entrepreneurship is not automatically a functioning and that valuable opportunities are not necessarily matched through their work. For self-employed workers, capability building is also more challenging by definition, as they mainly rely on individual rather than collective capabilities (Ibrahim, 2017) that regular workers can generate in the firm that employs them.

In this article, we investigate institutional differences in the emergence of work-related functionings between regular and self-employed workers (Table 1). Particularly, we aim to answer questions including: in which context do regular workers benefit from more institutional support than self-employed workers and why; how job stability and autonomy at work are shaped in each context; and finally, what happens in terms of access and participation to training programmes as well as institutional protection by health and safety and unions for both categories of workers. For this purpose, we exploit data from the European Working Conditions Survey (EWCS) and look at how theoretical nuances of Sen's (1985) capability approach are reflected in the Belgian, French and Dutch societies. Our aim is to shed light on how different institutional contexts can shape the respective functionings of self-employed workers and regular employees according to certain hypotheses. Namely, we expect a larger gap between the two categories of workers in terms of job instability in Belgium and France, which present less employment protection for self-employed workers (Hypothesis 1). We also expect a larger disadvantage for self-employed workers to partake in training in these countries, where contrarily to the Netherlands, active labour market policy is not consistent for them (Hypothesis 2). Finally, we believe the lack of social rights, collective bargaining and dialogue between labour market institutions in support of self-employed workers in Belgium and France to produce a larger gap between self-employed and regular workers in terms of institutional support (Hypothesis 3).

On this subject, a recent analysis by Psychologios et al. (2020) focuses on two historically different but regionally connected countries, namely, Greece and Serbia, with respect to their different institutionalisation of working conditions. In our article, we focus on the European triangle of Belgium, the Netherlands and France due to the contrasting institutional frameworks they present, especially regarding regulation and support of work —and specifically, self-employment. While both Belgium and the Netherlands are characterised by a centralised level of collective bargaining that allows for high coverage rate, differences exist with respect to the density and role of trade unions. Belgian unions, for instance, provide direct financial support to their unemployed members and are also more proactive in policymaking. Collective bargaining institutions in France are instead “uncoordinated so that they have a greater dependence on state regulation” (Beuker and Pichault, 2022). Such frameworks, then, have significant repercussions on self-employed workers. This category of workers is excluded from classical trade unions in both Belgium and France —where collective agreements are only possible for specific categories— and is instead included in specialised and inclusive unions in the Netherlands —also pioneers in representing solo self-employed workers.

In particular, in Belgium, the legislation regulating collective bargaining does not consider the self-employed category of workers as one directly able to set bargaining terms and conditions (Fulton, 2018). Self-employed workers are indeed considered employers rather than workers (law 5 December 1968, art. 3). On this subject, a specific classification test has been put in place to distinguish among categories of workers and avoid social fraud (Countouris and De Stefano, 2019). While self-employed workers are somewhat protected in terms of occupational health and safety norms, as well as discrimination, “the regulation governing and protecting collective bargaining does not apply to the collective bargaining of self-employed workers vis-à-vis their principals” (Countouris and De Stefano, 2019). Similarly, in France, collective bargaining for self-employed workers is not legally covered by the legislation. Some self-employed workers, however, may be counted as employees when they fall under certain categories; namely, journalists, actors and others working in the creative arts, “and those working autonomously, but who are not completely independent” (Fulton, 2018). In contrast, the Netherlands offers clear autonomous negotiating power to the self-employed workers since 1927. Their trade unions are strongly institutionalised in the industrial relations system (Vandaele and Leschke, 2010), which makes debates on work-related issues easier also for self-employed and non-standard workers.

According to recent EU statistics, the Dutch labour market presents a high-quality level of industrial relations, industrial democracy, social justice and general quality of work and employment. The Netherlands is also characterised by a lower rate of unemployment and a significantly higher flexibility in the labour market. Associated with this, the country registers a larger number of self-employed workers and, consequently, an open attitude and efficient proactivity of trade unions towards them, including investment in vocational training and worker's agency. Comparatively to the Netherlands, Belgium and France position themselves at quite the lower rank in institutional support for self-employed workers; Belgium as regards representation and participation rights and social dialogue at the firm level and France as regards the capability of labour market institutions to provide workers with the high-level means to achieve good career prospects and well-being. Among European countries, France remains above unemployment rates average, with labour market liberalisation still in progress and a stagnant low rate of self-employment for high-skilled workers. The skill level remains low also in Belgium, together with work incentives and lifelong learning practices. Social assistance for atypical workers has progressed over time. However, the rigid employment protection schemes do not allow for a flexible labour market overall. In line with the argument by Lefebvre et al. (2015), it is evident that these three countries “share the same social and fiscal institutions but not necessarily the same values and social norms”; thus, suggesting that outcomes as regards the building of capabilities in the labour market could differ. In this article, we investigate such differences by looking at the above-mentioned work-related functionings.

To achieve this, we exploit the most recent collection of observational data at the individual level; namely, the European Working Conditions Survey (EWCS). This was established in 1990 and since then represents a valid instrument for the EU Eurofound Agency to investigate the working conditions of workers in the European territory. The survey allows to assess the quality of employment, monitor groups at risk and identify physical and psychological issues in the world of work, to ultimately understand “the everyday reality of men and women at work”.

2. Literature review

In an extremely flexible labour market, employees and self-employed workers move from full-to part-time employment, as well as from employment to unemployment differently. If regular workers can rely on their employer, self-employed workers must usually rely on their skills (Menger, 2017). While the quality of self-employment in a country is often associated to the country's wealth or its investment in R&D (Burke et al., 2021), self-employed workers usually report higher levels of stress due to job instability (Bencsik and Chuluun, 2021). Below, we show that the range of functionings observed in the context of work can differ between regular and self-employed workers according to the institutional framework of reference. To do so, we first define such functionings according to the extant literature on working-life conditions and then identify them in the EWCS conducted in 2015.

Particularly, we define work-related functionings as those “enabling” factors that allow the individual to reach her full potential, or capability to do so, at work. In this category, we include the following variables [1]. After distinguishing between regular and self-employed workers, we investigate for how long the individual worked in the firm in accordance with the hypotheses brought forward by Kryscynski et al. (2021) that incentives developed in a specific firm “provide more utility to workers” and by Gartenberg and Zenger (2022) that firms work as subsocieties in which individuals identify. Secondly, we look at whether the worker experienced a change in salary or income in the past year, as well as whether she experienced a change in the working hours in the past year. This is justified by the idea that job instability not only influences work performance (Probst et al., 2017) and attachment to the labour market, but equally so the private and social spheres of an individual. In addition to affecting personality traits in the long term (Wu et al., 2020), recent analyses show that unstable employment favours divorce (Kaplan and Herbst-Debby, 2018), antiegalitarian attitudes in society (Selenko and De Witte, 2021), as well as poor physical and mental health (Minnotte and Yucel, 2018). On this subject, Ellorenco et al. (2019) recently demonstrated how wage distribution, and particularly, wage justice–often not accounted for in self-employment–directly affects life capabilities measured as life satisfaction and physical wellbeing. We hypothesise that a larger difference in hours worked and income between self-employed and regular workers will be registered in the Belgian and French institutional settings (Table 1).

With respect to factors linked to human capital investment, we identify whether in his work, the individual is able to change tasks, as well as methods; if the worker participated in regular training provided in the workplace or other forms of training. The data confirm differences between self-employed and regular workers overall. Regular workers may be more positively committed to on-the-job training measure, while self-employed workers rely more consistently on forms of training. Self-employed workers may have less access to learning opportunities due to these being potentially perceived as an obstacle to generation of revenues. In addition to financial constraints, self-employed workers may avoid participation in training due to short-termism: having time and project constraints they may not see the value in learning new skills. Self-employed workers may also be subject to taxation of self-funded work-related training. However, we expect this gap to be larger in countries where a non-supportive institutional framework towards self-employed workers does not allow for consistent active labour market policy for this category; namely, France and Belgium (Table 1). On the other hand, it is likely for self-employed workers to experience independence in change of tasks and methods more easily given the autonomous nature of their employment. Having said that, we expect for the Dutch institutional system to enhance such independence at work, while for the French and Belgian non-supportive institutional contexts to strengthen the gap in disfavour of self-employed workers as regards financial instability (Table 1).

Extant research shows that being flexible encourages initiative and self-efficacy (van den Berg and van der Velde, 2005). At the same time, changes in work content and procedures often require reskilling to be beneficial for workers. This is why we also investigate potential differences between regular and self-employed workers with respect to training. In addition to increasing productivity and quality of work, human capital investment also affects social outcomes such as interpersonal trust or participation in activities such as volunteering (Vera-Toscano et al., 2017). This is in line with the argument by Subramanian et al. (2013) that human capital should not be considered just from the viewpoint of economic development, but with regards to its human component as well. In other words, it should contribute to the feasibility of aspirations (Sarojini Hart, 2016). As skills empower people, they also enhance opportunities (Heckman and Corbin, 2016).

Finally, regarding the institutional support provided to the worker, we consider whether there is a health and safety organisation of reference; whether there is a trade union that represents them; and for regular workers, whether there is an organisation that discusses workers' views. With the limitations present in the EWCS survey, we also refer to worker's agency and empowerment by looking at whether the worker declares to have a direct power in influencing decisions that are important to his work.

This in line with the analysis by Pichault and McKeown (2019), which illustrates how the level of independence typical of non-standard employment may lead to low levels of worker protection. On this subject, it is also relevant to think of the differences observed in Belgium, France and the Netherlands with respect to trade unions' density and roles. Vo-Thahn et al. (2022) have recently shown that trade union support reduces emotional exhaustion in times of job insecurity like the current ones, while Li et al. (2019) depicted the benefits related to work associated with union participation —including stronger job involvement and higher levels of organisational justice. Understanding these dimensions is, therefore, crucial to shed light on eventual gaps antagonistic to self-employed workers. The confusion present in the Belgian and French legislation as regards the definition and categorisation of self-employed workers into regular employees or regular employers makes it difficult for the appropriate institutions to step in and guarantee occupational safety, as well as empowerment of the individual's agency and negotiating power to self-employed workers. In contrast, in the Netherlands, the more flexible but ad hoc legislation in support of self-employed workers allows for proper institutional support, which makes us hypothesize that, in this country, the gap between self-employed and regular workers is smaller (Table 1).

3. Data and methods

For the purpose of our analysis, we create variables that proxy various sub-categories of work-related functioning based on primary data. These refer to 5,142 individuals working in Belgium, the Netherlands and France and interviewed by officials of the EWCS in the year of 2015. Information is available with respect to their individual characteristics, such as gender and age; as well as their working conditions, including the type of job they carry out –although it is not possible to distinguish between voluntary and involuntary self-employment–, the type of firm that employs them, the salary they earn, the hours they work (Tables 2–5). Proxies for work-related functionings are then reflected in various sub-categories: (1) job stability –namely, change experienced in hours worked, change experienced in income, possibility to change tasks and methods–; (2) human capital investment –or participation in forms of training; and (3) institutional support –namely, relationship with a trade union, health organisation, as well as worker's agency. These are all defined as dummy variables. We also provide estimates from a regression analysis that accounts for individual characteristics at the country level. For this purpose, we also include information on the aggregate unemployment rate registered in Belgium, France and the Netherlands (see Table 5).

Regarding the Belgium sample, 50.1% of the workers interviewed are men and 49.9% are women —despite the gender wage gap, female labour participation has increased over the years. The average age for such workers is 43.2 years. These patterns are also observed for France and the Netherlands, with a slightly higher proportion of women; 52.4 and 50.2%, respectively. When investigating work-related characteristics, we find that at least 83.8% of the Belgian individuals interviewed by EWCS officials were employed in firms in 2015, while the remaining 17.2% were self-employed. On average, individuals have been working at the same firm for 11.7 years. In France, the proportion of self-employed individuals is even lower (7.9%), while slightly more people experienced a decrease in their salary in the past twelve months (11.3%). On the other hand, the situation differs in the Netherlands. Here, 17.1% of the individuals are self-employed, while people tend to work for fewer years in the same firm (11 years, on average). The slightly higher tendency for Dutch workers to opt for self-employment may be due to a lower differential between self- and regular employment (De Wit and Van Winden, 1989), tax incentives in favour of entrepreneurship offered by the government (Van Andel and Loots, 2021), as well as the presence of many organisations, other than trade unions, providing ad hoc support to self-employed workers. The larger inclusion of self-employed workers in systems of active labour market policy based on reskilling, as well as in welfare systems may make it more attractive for individuals to opt for self-employment as a voluntary and positive form of employment.

On this subject, Beuker and Pichault (2022) recently showed how the institutional context is crucial in shaping industrial relations, as well as the modalities with which certain categories of workers are represented in the labour market. Job satisfaction is, indeed, significantly influenced by both union membership and employees' empowerment (Van der Meer, 2019). While in both Belgium and the Netherlands social partners are highly involved in policymaking—according to what is usually referred to as institutional partnership—, in France, the collective bargaining institutions are quite uncoordinated and mostly regulated at firms and industry-levels. On this subject, an article of Pulignano (2018) explains how such decentralisation allows companies “to reach deals of their own with workers rather than being forced to comply with multi-employer industry-wide agreements negotiated by representative trade unions”. At the same time, differences exist among the three countries with respect to the rate of unionisation —the Belgian and the French ranking, respectively, highest and lowest—, or the inclusion of self-employed workers in unions. In the latter case, the Netherlands stands out as providing the most flexible and hybrid solutions to the self-employed population (Beuker and Pichault, 2022). In the sections below, we illustrate potential differences present in the three countries regarding job stability, human capital investment and institutional support for regular versus self-employed workers.

4. Results

4.1 Job stability

Recent analyses by Lübke and Erlinghagen (2014) show that self-perceived job insecurity varies across European countries, in terms of both cognitive job insecurity —intended as perceived likelihood of job loss— and labour market insecurity —intended as self-perceived difficulties of job search. Between 2004 and 2010, for instance, change in perceived job insecurity was equal to 0.2, −2.4 and −1.2 in, respectively, the Netherlands, France and Belgium. While, overall, this suggests that Dutch, French and Belgian workers are quite confident that they are unlikely to become unemployment, the situation differs when it comes to their perception of job search. In both the Netherlands and Belgium, workers report lower perceived difficulties in job search (−3.8 and −5.4), while in France we register an increase in the same rate (+1). This, then, connects to the institutional context that countries offer in terms of employment protection, also with respect to self-employment. In general, self-employed workers are affected by changes in payment systems offered by used contractors and platforms, the taking up of new jobs, the introduction of subsidies, or the variable change in demand by clients. These are all factors that could directly affect working hours (Figures 1 and 2).

In the Dutch case, for instance, flexicurity policies allow for a normalisation of atypical work, especially for the female labour force (Bekker and Wilthagen, 2008), supported by unemployment benefits, on the one hand and reskilling policies, on the other (Keune, 2008). This contrasts with France, characterised by a rigid labour market, in terms of both wage adjustment and job mobility (Vlandas, 2017). Similarly, people's careers in Belgium are usually affected by low dynamism, while assistance lacks with respect to job-to-job transitions (McGowan et al., 2020). We are interested in factors that are associated with stability in the job performed by the individual, including whether the worker has experienced a decrease in salary or hours worked, in the past year, as well as whether she is able to change tasks or methods during her job.

When looking at the Belgian case, we observe that the proportion of self-employed whose income dropped in the past twelve months (14.6%) is more than twice of that of regular workers who experienced the same decrease (5.3%). The contrast is similar when looking at negative changes in hours worked. 11.7% of self-employed workers reported a decrease in working hours, compared to just 6.1% of regular workers. This is relevant once we know that atypical forms of employment, including precarious work and self-employment, significantly and negatively affect individuals' mental health in Belgium (Van Aerden et al., 2017). This can be explained by factors such as lower financial resources, scarcer social networks, higher autonomy in time disposal and negative self-esteem mechanisms. On the other hand, the comparative advantage usually attributed to self-employed workers in terms of ability to change the ways in which they work, is relatively small in Belgium. The data, in this case, show that 83.5 and 71.2% of, respectively, self-employed and regular workers were able to change tasks in the past year. The difference increases slightly with respect to ability, or capability, to change methods performing their job. The proportion was equal to 84.2% for self-employed individuals and 71.7% for regular workers. This is relevant as changes in the job contribute to enhancing career development (Chen et al., 2004) and, therefore, increase the individual's occupational and consequently personal, well-being.

Regression results confirm our hypotheses. Being a regular worker decreases the chances of experiencing a drop in income and working hours by, respectively, 11.0 and 7.0% points, significant at 1% level. In this regard, female workers experience a slightly larger likelihood of working less hours over the long run (1.0). On the other hand, regular workers are 16.9 and 18.0% points less likely to, respectively, be able to change the order of the tasks carried out and methods used at work (see Tables 6–9).

In France, the stability experienced by self-employed individuals in terms of salary and hours of work is critically different compared to regular workers. Not only are self-employed individuals who experienced a drop in number of hours worked in the last twelve months (19.1%) three times the size of regular workers with a similar occupational drawback (6.8%). Their earnings also decreased significantly. While only 9.1% of workers in regular employment report a cut in their salary, in the case of the self-employed at least 33.1% did. Results from a regression analysis by country show that regular workers are indeed 23.4 and 11.7% points less likely to experience, respectively, a drop in income and in working hours. This connects to the analysis by Georgieff and Lepinteur (2018), which shows how the intervention of the French government, in the form of negative incentives applied to firms when they lay off workers, increased perceived job security but only for certain categories of workers. Finally, while the proportion of French regular and self-employed capable of changing tasks in their work is similar to that of Belgian workers, the divergence increases regarding ability to change methods. In France, regular workers are significantly less likely to do a job where they can apply different methods (68.5%), compared to self-employed workers (86%). In particular, being a regular worker, as opposed to a self-employed worker, decreases this autonomy by 18.1% points, significant at 1% level. The degree of autonomy in work organisation increases only slightly for workers committed to the same firm with each additional year (0.2). This is not surprising as French labour law explicitly values seniority (Chatzilaou, 2022).

The Dutch case outstands in that the difference in the decrease of hours worked between regular and self-employed workers experience is, on average, quite low (2.1%). While we cannot retrieve information as regards the reasons why self-employed workers experienced reduction in wages and hours, what is observed in the Dutch labour market is possibly explained by the fact that employees, in this country, have no problem in switching from full-to part-time employment, which in 2015 accounted for 56.6% of total employment —a significantly higher proportion than what is observed in the two other countries under investigation. On the other hand, the trend relative to decreased earnings simulates the one observed for French workers, with 18.8 and 9.6% of, respectively, self-employed and regular workers having experienced a decrease in their earnings, in the past twelve months. However, the explanatory effect is slightly smaller than was observed in Belgium and France. Being a regular worker decreases the likelihood of experiencing a cut in income by 9.2% points.

Regarding Dutch workers' capability of changing tasks and methods, we observe that the proportion of self-employed “freedom” to change tasks and methods is the highest, compared to what happens in France and Belgium. In the Netherlands, 91.5 and 89.7% of self-employed workers report being able to, respectively, change methods in their work and carry out different tasks —rates for regular workers are, instead, equal to 70.1 and 75.1%. The effect, in this case, is comparatively larger in favour of self-employed workers. Being a regular worker decreases the capability of changing methods at work autonomously by 21.7% points, significant at 1% level. Older workers, in general, are 0.3% points more likely to experience independence in changing the type and order of tasks carried out at work. The fact that the Dutch government guarantees protection to those who choose a more flexible type of job may explain the larger share of own-accounted self-employment in the country (Schulze-Buschoff and Schmidt, 2009).

In general, we observe larger effects for the Netherlands as regards flexibility in work organisation, while effects are more economically significant for Belgium and France as regards the divergence between regular and self-employed workers in maintaining financial stability over time. This is in line with our Hypothesis 1 that a less supportive institutional framework towards self-employed workers enhances the disadvantage of this category for functionings related to job stability, reducing their quality of work (Table 1). The country's unemployment rate affects these results only slightly. Where the unemployment rate is higher, the drop in income is 0.6% points more likely to occur, significant at 10% level.

4.2 Human capital investment

Apart from increasing the well-being of nations by creating forms of cognitive capitalism (Rindermann, 2018), human capital investment significantly influences individuals at the personal level too. On the one hand, human capital helps the individual develop psychological characteristics that are fundamental for creating a positive experience in the workplace (Okun, 2022). On the other hand, by feeling good at work, either due to accomplishments or social relations, individuals tend to feel happier also at home (Rodriguez-Munoz et al., 2014). While the EWCS survey does not capture the quality of training, we are still interested in understanding how the distribution of, or participation in, training for work purposes varies between regular and self-employed workers in the countries under investigation. On this subject, Lecourt (2013) underlined how it is more the socio-economic issues at stake, rather than the individual characteristics, to shape capability pathways. When it comes to public spending on labour markets, for instance, Belgium, France and the Netherlands are among the most generous OECD countries. In 2014, their governments spent, respectively, 2.390, 2.970 and 2.550 as a percentage of their GDP in labour market policies. Differences, however, may emerge when looking at regular and self-employment (Figure 3).

In Belgium, we observe that 42.5% of regular workers participated in on-the-job training. On the contrary, only 12% of self-employed workers did. The latter are more likely than regular workers to participate in other forms of training, also self-provided (12% against 10.5%). Results from a regression analysis show that regular workers are 28.2% points more likely to take part in on-the-job training, significant at 1% level, with slightly higher opportunities for senior workers (0.4). No significant effect is registered for participation in alternative forms of training. Even though Belgium benefits from a history of large public spending in active labour market policies (Grubb and Martin, 2001) —some of them dedicated to reskilling—, it is relevant to stress that Belgian atypical workers get fewer opportunities from training compared to permanent employees (Forrier and Sels, 2003). This is in line with the argument by Rueda (2006) that active labour market policies like training usually benefits those with secure employment rather than “outsiders” such as self-employed workers. The fact that self-employed workers must invest in training using their own sources is an evident disincentive for them to take part in them; therefore, decreasing the set of occupational opportunities that may derive from it. In general, active labour market policy in the domain of self-employment is often reflected in start-up incentives (Román et al., 2013) that promote entrepreneurship rather than human capital investment that could enhance individual capabilities.

The French case presents similarities with Belgium in that 32.9% of regular workers participated in on-the-job training, while only 11.7% of self-employed workers took part in it. On the other hand, self-employed workers participate in alternative forms of training more often. In particular, regular workers are 21.2% points more likely to take part in regular forms of training and 8.3% points less likely to take part in other training, instead popular in self-employment. This is not surprising when we note that, in general, only around 10% of firms in France are defined as “capability-friendly” (Lambert and Vero, 2013). This means they are considered guarantors of training policies that enhance individuals' aspiration to learn further, creating a so-called “appetence for learning” (Lambert and Vero, 2013).

In the Netherlands, on the other hand, more attention seems to be given to training. We observe 43.9% of regular workers as having participated in training provided by their employer in the past year, as well as 15.8% of self-employed workers having taken part in alternative training. While no significant results are observed for alternative training, due to the small number of workers participating in it, our regression results show that the gap between regular and self-employed workers is statistically and economically significant when it comes to on-the-job training. In this case, regular workers have an advantage of 34.5% points. The conspicuous amount of public money spent every year in active labour market policy in the Netherlands (Lammers and Kok, 2021) allows self-employed individuals to benefit from reskilling to a larger extent than what happens in Belgium and France. Because “costs for training and education can be accounted as business costs and are thus tax deductible” in the Netherlands (Bekker, 2010), self-employed individuals may have stronger incentives to take part in alternative reskilling programmes.

Overall, regular workers benefit from training measures more consistently than self-employed workers. This is observed in all three countries under investigation especially as regards regular training. In countries with a higher unemployment rate, workers are 2.9% points less likely to participate in training, irrespective of the type of employment. Consistently with our Hypothesis 2, we observe a lower gap between self-employed and regular workers in access to training in the Netherlands, institutionally characterised by a more efficient, ad hoc, and generous system of active labour market policy (Table 1).

4.3 Institutional support

In addition to varying across Europe (Keune and Pedaci, 2020), trade union strategies and trade union density differ significantly in the European triangle considered in this analysis. With respect to the latter, this was equal to 52.9% in 2014 in Belgium, followed by the Netherlands with 18.1% and France with only 11%. Being supported by the right labour market institutions can help workers to both develop their skills, increasing their functionings at work and protect their rights, potentially increasing their functionings in the private sphere. In addition to high-involvement management practices, workplace performance usually improves when there is trade union representation (Bryson et al., 2005). Similarly, the presence of a health and safety management system in the firm is generally associated with higher levels of perceived support from management and colleagues, as well as of engagement in health and safety activities (Torp and Moen, 2006; Walters and Wadsworth, 2020). Other labour market actors may engage with self-employed workers, but these are usually not represented in official aggregate statistics, as is the case for the EWCS survey (Figures 4 and 5).

The divergence between regular and self-employed workers with respect to exposure to institutions that ameliorate working conditions is evident in all three countries under investigation. For Belgium, only 1.2% of self-employed workers, against 64% of employees, is referred to a health and safety institution when working for themselves. Looking at regression results, this gap is confirmed. Regular workers are 65.6% points more likely than self-employed worker to be protected by a health and safety institution. The same trend is observed as regards the presence of an organisation that discusses workers' views (55.4). Conversely, self-employed workers (93%) are advantaged compared to regular workers (69.2%) with respect to the power experienced in influencing decisions that affect their work. Regular workers are 23.7% points, significant at 1% level, less likely to do so. However, relatively to trade union membership, we observe that 63.5% of Belgian regular workers are employed in a firm represented by a trade union, while only 1.4% of self-employed workers is associated to one.The effect, in this case, is statistically and economically significant and equal to a 64.3% points advantage compared to their self-employed colleagues. In general, the high rate of union membership present in the country is typical of a neo-corporatist regime where social partners are more active than the government itself in influencing collective bargaining, but social rights for the self-employed remain low compared to regular workers or workers in the public sector (Beuker et al., 2019). In other words, strong protections in favour of regular workers may threaten work opportunities for certain categories of workers, including the youths, women and migrants, who may involuntarily end up in the less protected self-employment.

Self-employed individuals follow a similar pattern in France. On the one hand, French firms seem to provide support of the abovementioned type to the same extent to regular workers —this is not surprising given the high rate of strikes present in the country and the primary role attributed to social partners (Gazier, 2019). 59.7% of individuals in this category can be referred to a health and safety representative in their firm, 56.5% of them have the possibility to express their view in the company, and 0.62.5% of employees work in a firm supported by a trade union Looking at our regression results, we observe that regular workers are 62.2% points more likely to be supported by a trade union. Female and older workers are less likely to receive institutional support, especially as regards health (−9.4 and −0.1, respectively). In parallel, 96.1% of self-employed workers claim they have power in influencing decisions related to their work. On the other hand, self-employed workers are completely excluded from such institutional support. This is in line with the regulatory framework in place, which –in general– does not encourage the enhancement of workers' capability set (Bonvin et al., 2013).

In French labour law, “self-employed workers are defined by what they are not: neither employee nor agricultural worker” (Beuker et al., 2019). While both the labour code and commercial code provide legislative ad hoc measures, distinguishing between self-employed workers, auto-entrepreneurs and hybrid forms of self-employment, public policy support is still unequal towards this category. As justly observed by Beuker et al. (2019), self-employed workers are subject to benefits insofar they belong to categories not necessarily related to their occupation; namely, they receive support if they become parents, undergo unemployment, reach old age and so on. In other words, the French welfare system does not treat self-employment as an occupational choice deserving of financial or institutional support, unless related to the issues mentioned above.

The institutional framework observed in the Netherlands presents similarities with the Belgian one. For the year 2015, we observe 59.7% of regular workers as being employed in a firm associated with a trade union, while the proportion drops to 0.6% for self-employed workers. Attention to non-standard workers in the Netherlands has significantly increased since 2010, when the self-employed have obtained a seat in the Dutch Social and Economic Council. Nevertheless, the concept of self-employment remains vague in the country. The comparative advantage of regular workers in the Netherlands appears somewhat smaller as regards trade union support but it is still economically significant (59.7% points) and influenced by seniority (0.9). While there exist organisations that define themselves as “associations of self-employed, for self-employed, by self-employed”, self-employed workers like project-based workers may also be treated as quasi-employees who respond to specific employers (Pichault and McKeown, 2019). Thus, on the one hand, some independent workers may end up being represented by the more standard trade unions. On the other hand, some self-employed may register with solo self-employment organisations such as the ZZPs (Jansen, 2017), not officially counting as trade unions.

The observed differences between regular and self-employed workers all seem to converge to a welfare system very protective of regular employment and dubious of the necessary steps to take with respect to “atypical” jobs that have become more and more typical. While French employees are extensively covered by collective bargaining, the outsiders who are left out —namely, the self-employed— find themselves wandering in a legislative framework that remains abstract for what concerns their rights as workers. In other words, in both France and Belgium, work-related functionings that may emerge from receiving the just institutional support remain a privilege appreciated by regular employees. Conversely, self-employed workers in the Netherlands are not dependant on special authorisations for protection of their social rights. The unconventional inclusivity that characterises the Dutch case implies self-employment receives similar, despite not yet sufficient, attention by social partners. As a consequence, independent workers are provided with the institutional instruments that allow them to develop their full potential at work and outside of it.

Overall, our findings confirm our Hypothesis 3. The institutional framework in each country was a determinant of the stronger or weaker gap between self-employed and regular employees as regards the support of a trade union, occupational safety and workers' agency. In particular, the situation of self-employed workers appears to be less detrimental in the Netherlands due to a more inclusive attention by social partners and supportive institutional framework that protects self-employed workers through ad hoc regulations, social rights and organisations (Table 1).

5. Conclusions

Building on Sen's theoretical framework (1985), in this article, we defined work-related functionings as those factors relative to work that facilitate the individual's possibility to reach his/her full potential at work. Based on primary data referring to 5,142 individuals interviewed via the EWCS, we created proxies for work-related functionings. Particularly, we contributed to the extant literature by shedding new light on potential differences between regular and self-employed workers in Belgium, France and the Netherlands. We hypothesised and showed that where institutional contexts are non-supportive towards self-employed workers, there exists a higher gap in work-related functionings for self-employed workers, compared to regular workers and that this eventually leads to a lower probability of good quality of work for self-employed workers.

We found that the three countries under investigation diverge with respect to providing workers with equal “opportunities”, or work-related functionings, depending on the category of employment they associate with. In both Belgium and France, self-employed workers are subject to higher instability in terms of changes in salary and hours worked compared to the Netherlands, where atypical work is better positioned in the labour market. The Dutch case also stands out when it comes to exposing workers to training. While the gap between participation in training between regular and self-employed is still positively biased in favour of those who work in regular employment, a further gap emerges between self-employed workers in Belgium, France and the Netherlands. The Dutch labour market seems to be more inclusive with respect to allowing self-employed workers to develop their work-related functionings from a human capital perspective too.

Our research presents several limitations. First, it focuses specifically on a well-developed European area on a topic that is dependent on the institutional framework present in the country. This means that results may differ in institutionally different European regions, including the Southern countries of Italy or Greece. Second, our article highly depends on data limited to 2015. For this reason, it is important to note that results may vary when considering more distant periods in time —for instance, the 2008 economic crisis—, or the more recent pandemic years of economic stagnation and institutional difficulties. Third, we focus on regular and self-employed workers in general, not distinguishing between the extant economic categories —from individuals working in agriculture, industry, services.

Future research could exploit data on working conditions that account for the recent Covid-19 pandemic and its impact on different categories of workers, especially as regards working patterns (e.g. the involuntary transition of some into non-standard work) and working hours (e.g. the trade-off between autonomy and collapse of leisure time). On this subject, new research could investigate the meaning of other functionings related to skill formation, the pension system and flexibility in time dedicated to work. Similarly, within-category divergences could be explored as regards the extant models of self-employment –from dependent and independent self-employed workers to self-employed workers with employees and solo traders–, where risks could vary both institutionally and industrially. This, in turn, could give origin to new debates on capability building based on the different nature of trade-offs experiences by self-employed workers. For this purpose, datasets other than the EWCS should be used to understand perceived individual job quality. Our analysis presents several implications for society and policymaking. On the one hand, we show that self-employed workers encounter greater difficulties in developing their full potential in the world of work. This confirms the findings from previous research. Governments in this European triangle indirectly encourage a model of flexinsecurity, where regular workers benefit from financial stability and paid training but not enough work flexibility, while self-employed workers trade off job stability and institutional protection for independence in their management of work. As a consequence, a combination of passive (based on financial support) and active (based on human capital) measures should be provided to regular and self-employed workers in ad hoc form. The nature of their work and type of institutional protection widely differs and hence requires labour market policies that are different in content and temporal characteristics (e.g. duration, flexibility) but equal in efficacy.

On the other hand, our findings shed light on the inability of governments to provide self-employed workers with a welfare system that goes beyond workfare. Providing individuals with the right functionings at work, in the guise of workers, counts for their possibility to develop the functionings they care about outside of work, as individuals. Knowing that work-related functionings significantly impact social functionings as well, the inequality between regular and self-employed workers becomes even more worrisome. For years, the labour market exploited utility models associated with the professional survival and protection of those able to produce extensively and fast and conform to “standard” jobs, ignoring the real talents or necessities of individuals as people. We recommend economic agents such as the state, trade unions and employers to provide individuals with instruments at work that encourage the building of educational, financial and societal capabilities also outside of work. The provision of portable social rights, independent from the work status, already implemented in Sweden for a long time, is a concrete way to progress in this perspective (Pichault and McKeown, 2019).

According to Sen (1985), the success of a society is to be measured by the freedoms that individuals in that society enjoy. The fact that regular workers represent most of the working population in Europe should not make self-employed workers, and other workers on non-standard contracts, less worthy of such freedoms.

Figures

Figure 1

Experienced decrease in income and hours worked in the past twelve months

Figure 2

Ability to change order of tasks and methods

Figure 3

Participation in on-the-job training and alternative forms of training

Figure 4

Active presence of trade union and health and safety organisation

Figure 5

Active presence of worker's organisation and worker's agency

Table 1

Institutional influences on work-related functionings for self-employed workers (compared to regular workers)

Note(s): ALMP is active labour market policy, F is France, B is Belgium, N is Netherlands and SEW is self-employed workers

Table 2

Descriptive statistics for workers interviewed by EWCS officials in Belgium, France, and The Netherlands in 2015

Mean (%)	Belgium	France	The Netherlands
Demographics characteristics
Female = 1	49.9 (50)	52.4 (50.0)	50.2 (50.0)
Age (years)	43.2 (12.1)	43.7 (27.0)	44.4 (13.7)
Employment characteristics
Employee = 1	83.8 (37.0)	91.1 (28.0)	83.9 (37.0)
Years in firm	11.7 (11.1)	11.3 (10.4)	11.0 (10.8)
Decreased salary = 1	6.7 (25.0)	11.3 (32.0)	11.1 (31.0)
Decreased hours = 1	7.0 (26.0)	7.9 (27.0)	12.2 (33.0)
Able to change tasks = 1	73.9 (44.0)	73.8 (44.0)	77.4 (42.0)
Able to change methods = 1	73.7 (44.0)	70.1 (46.0)	73.5 (44.0)
Training
On-the-job training = 1	37.5 (48.4)	30.9 (46.3)	38.1 (48.6)
Other forms of training = 1	10.7 (30.9)	11.8 (32.3)	14.2 (34.9)
Institutional support
Trade union = 1	53.5 (49.9)	56.9 (49.5)	50.2 (50.0)
Employees' views = 1	48.1 (50.0)	51.5 (50.0)	49.1 (50.0)
Health and safety organisation = 1	45.7 (49.9)	49.8 (49.8)	45.7 (49.8)
Worker's agency = 1	72.5 (44.7)	71.1 (45.4)	82.5 (37.9)
Observations	2,587	1,527	1,028

Note(s): The table shows descriptive statistics on the working conditions of employees and self-employed individuals interviewed in Belgium, France, and The Netherlands by EWCS officials in 2015. Information on training and change in factors linked to employment refer to the past 12 months. Standard deviations in parentheses

Table 3

Regular and self-employed workers in Belgium (2015)

Work-related functionings	Regular workers (%)	Self-employed workers (%)
Job stability
Decreased salary = 1	5.3	14.6
Decreased hours = 1	6.1	11.7
Able to change tasks = 1	71.2	83.5
Able to change methods = 1	71.7	84.2
Human capital investment
On-the-job training = 1	42.5	12.0
Other forms of training = 1	10.5	12.0
Institutional support
Trade union = 1	63.5	1.4
Employees' views = 1	57.1	1.7
Health and safety organisation = 1	64.0	1.2
Worker's agency = 1	69.2	93.0
Observations	2,169	418

Note(s): The table shows differences between regular and self-employed workers with respect to work-related functionings observed in Belgium in 2015

Table 4

Regular and self-employed workers in France (2015)

Work-related functionings	Regular workers (%)	Self-employed workers (%)
Job stability
Decreased salary = 1	9.1	33.1
Decreased hours = 1	6.8	19.1
Able to change tasks = 1	72.5	87.5
Able to change methods = 1	68.5	86.0
Human capital investment
On-the-job training = 1	32.9	11.7
Other forms of training = 1	11.1	19.1
Institutional support
Trade union = 1	62.5	0
Employees' views = 1	56.5	0
Health and safety organisation = 1	59.7	0.7
Worker's agency = 1	68.8	96.1
Observations	1,391	136

Note(s): The table shows differences between regular and self-employed workers with respect to work-related functionings observed in France in 2015

Table 5

Regular and self-employed workers in the Netherlands (2015)

Work-related functionings	Regular workers (%)	Self-employed workers (%)
Job stability
Decreased salary = 1	9.6	18.8
Decreased hours = 1	11.8	13.9
Able to change tasks = 1	75.1	89.7
Able to change methods = 1	70.1	91.5
Human capital investment
On-the-job training = 1	43.9	7.9
Other forms of training = 1	13.9	15.8
Institutional support
Trade union = 1	59.7	0.6
Employees' views = 1	58.3	1.2
Health and safety organisation = 1	54.3	0.6
Worker's agency = 1	80.3	94.8
Observations	863	165

Note(s): The table shows differences between regular and self-employed workers with respect to work-related functionings observed in the Netherlands in 2015

Table 6

Regression results for building work-related functionings in Belgium (2015)

Y	On-the-job training	Other training	Decrease in hours	Decrease in income	Change tasks	Change methods	Trade union	Health and safety	Worker's organisation	Worker's agency
Regular worker = 1	0.282*** (0.277)	−0.025 (0.018)	−0.070*** (0.015)	−0.110*** (0.015)	−0.169*** (0.025)	−0.180*** (0.025)	0.643*** (0.026)	0.656*** (0.026)	0.554*** (0.027)	−0.237*** (0.026)
Female = 1	−0.011 (0.019)	−0.008 (0.012)	0.032** (0.010)	0.010* (0.010)	0.059*** (0.017)	0.043** (0.017)	−0.015 (0.018)	−0.026 (0.017)	0.019 (0.018)	−0.026 (0.018)
Age (years)	−0.004*** (0.001)	−0.001 (0.001)	0.002*** (0.001)	0.001 (0.001)	0.001 (0.000)	0.000 (0.000)	−0.002** (0.001)	−0.001 (0.001)	−0.001 (0.001)	−0.001 (0.001)
Years in firm	0.004*** (0.001)	−0.001 (0.001)	−0.000 (0.001)	0.001 (0.001)	0.002* (0.001)	0.001 (0.001)	0.010*** (0.001)	0.001*** (0.000)	0.004*** (0.001)	0.003** (0.001)
Cons	0.269*** (0.049)	0.098*** (0.032)	0.042 (0.026)	0.111*** (0.026)	0.781*** (0.045)	0.862*** (0.045)	−0.029 (0.045)	−0.052 (0.046)	0.027 (0.047)	0.949*** (0.046)
Obs	2,514	2,514	2,514	2,514	2,514	2,514	2,514	2,514	2,514	2,514

Table 7

Regression results for building work-related functionings in France (2015)

Y	On-the-job training	Other training	Decrease in hours	Decrease in income	Change tasks	Change methods	Trade union	Health and safety	Worker's organisation	Worker's agency
Regular worker = 1	0.212*** (0.042)	−0.083** (0.029)	−0.117*** (0.024)	−0.234*** (0.028)	−0.163*** (0.040)	−0.181*** (0.041	0.622*** (0.041)	0.590*** (0.041)	0.563*** (0.043)	−0.274*** (0.041)
Female = 1	−0.041* (0.024)	0.009 (0.017)	0.001 (0.014)	0.017 (0.016)	0.050** (0.023)	0.013 (0.023	−0.043* (0.023)	−0.094*** (0.024)	−0.042* (0.024)	−0.032 (0.023)
Age (years)	−0.001** (0.000)	−0.000 (0.000)	−0.000 (0.000)	0.001** (0.000)	−0.001** (0.000)	−0.001 (0.000)	−0.000 (0.000)	−0.001** (0.001)	−0.001** (0.000)	−0.001 (0.000)
Years in firm	0.001 (0.001)	−0.000 (0.001)	−0.001 (0.001)	0.000 (0.001)	0.003** (0.001)	0.002** (0.001)	0.008*** (0.0001)	0.001*** (0.0001)	0.005*** (0.001)	0.004*** (0.001)
Cons	0.174*** (0.046)	0.208*** (0.032)	0.193*** (0.027)	0.282*** (0.031)	0.862*** (0.044)	0.857** (0.046)	−0.054 (0.046)	−0.010 (0.046	0.012 (0.047)	0.957*** (0.046)
Obs	1,519	1,519	1,519	1,519	1,519	1,519	1,519	1,519	1,519	1,519

Table 8

Regression results for building work-related functionings in the Netherlands (2015)

Y	On-the-job training	Other training	Decrease in hours	Decrease in income	Change tasks	Change methods	Trade union	Health and safety	Worker's organisation	Worker's agency
Regular worker = 1	0.345*** (0.041)	−0.020 (0.030)	−0.026 (0.028)	−0.092*** (0.027)	−0.135*** (0.036)	−0.217*** (0.037)	0.597*** (0.038)	0.550*** (0.389)	0.578*** (0.039)	−0.141*** (0.033)
Female = 1	0.008 (0.029)	−0.019 (0.022)	0.058** (0.021)	0.022 (0.020)	−0.016 (0.026)	−0.074** (0.027)	0.043 (0.028)	0.008 (0.028	0.018 (0.028)	−0.020 (0.024)
Age (years)	−0.003** (0.001)	−0.000 (0.001)	0.000 (0.000)	0.001 (0.001)	0.003** (0.001)	−0.001 (0.001)	0.001 (0.001)	0.002 (0.001)	0.001 (0.001)	0.002 (0.001)
Years in firm	0.003* (0.002)	0.000 (0.001)	0.001 (0.001)	−0.000 (0.001)	0.001 (0.001)	0.003** (0.001)	0.009*** (0.001)	0.008*** (0.002)	0.005*** (0.002)	0.001 (0.001)
Cons	0.178** (0.067)	0.177*** (0.050)	0.109** (0.047)	0.150*** (0.045)	0.773*** (0.059)	0.953*** (0.062)	−0.148** (0.063)	−0.167*** (0.064)	−0.109* (0.065)	0.877*** (0.055)
Obs	1,026	1,026	1,026	1,026	1,026	1,026	1,026	1,026	1,026	1,026

Table 9

Regression results for building work-related functionings, overall

Y	On-the-job training	Other training	Decrease in hours	Decrease in income	Change tasks	Change methods	Trade union	Health and safety	Worker's organisation	Worker's agency
Regular worker = 1	0.284*** (0.020)	−0.038** (0.014)	−0.073*** (0.012)	−0.132*** (0.012)	−0.163*** (0.018)	−0.188*** (0.019)	0.628*** (0.019)	0.612*** (0.019)	0.562*** (0.020)	−0.220 *** (0.019)
Female = 1	−0.017 (0.013)	−0.005 (0.009)	0.028*** (0.008)	0.018** (0.008)	0.042*** (0.012)	0.010 (0.012)	0.012 (0.013)	−0.040** (0.013)	−0.000 (0.013)	−0.027** (0.012)
Age (years)	−0.002*** (0.000)	−0.000 (0.000)	0.000 (0.000)	0.001*** (0.000)	−0.000 (0.000)	−0.000 (0.000)	−0.000 (0.001)	−0.001** (0.001)	−0.001** (0.000)	−0.000 (0.000)
Years in firm	0.002*** (0.001)	−0.000 (0.000)	0.000 (0.000)	0.000 (0.000)	0.003*** (0.001)	0.002*** (0.000)	0.009*** (0.001)	0.009*** (0.002)	0.004*** (0.001)	0.003*** (0.001)
Unemployment rate	−0.029*** (0.005)	−0.004 (0.004)	−0.009** (0.003)	0.006* (0.003)	−0.006 (0.005)	−0.007 (0.005)	0.005 (0.005)	0.008 (0.005)	−0.005 (0.005)	−0.024*** (0.005)
Cons	0.425*** (0.052)	0.196*** (0.036)	0.197*** (0.031)	0.103*** (0.032)	0.893*** (0.048)	0.956*** (0.049)	−0.117** (0.049)	−0.120** (0.050)	−0.038 (0.051)	1.142*** (0.048)
Obs	5,059	5,059	5,059	5,059	5,059	5,059	5,059	5,059	5,059	4,945

Note

1. Note that, due to constraints from the dataset, other non-listed variables may be included in the category of work-related capabilities.