OPEN ACCESS I SSN 2280-4056 E-Journal of International and Comparative LABOUR STUDIES Volume 11 No. 03/2022 E-Journal of International and Comparative LABOUR STUDIES ADAPT International School of Higher Education in Labour and Industrial Relations Managing Editor Valeria Filì (University of Udine) Board of Directors Alexis Bugada (Aix-Marseille University), Valeria Filì (University of Udine), Anthony Forsyth (RMIT University), József Hajdu (University of Szeged), Shinya Ouchi (Kobe University), Daiva Petrylaite (Vilnius University), Valeria Pulignano (KU Leuven University), Michele Tiraboschi (Founding Editor - University of Modena and Reggio Emilia), Anja Zbyszewska (Carleton University). Editorial Board Labour Law: Emanuele Dagnino (University of Modena and Reggio Emilia); Tammy Katsabian (College of Management Academic Studies); Attila Kun (Károli Gáspár University); Adrian Todolì (University of Valencia); Caroline Vanuls (Aix-Marseille University). Industrial Relations: Valentina Franca (University of Ljubljana); Giuseppe Antonio Recchia (University of Bari Aldo Moro); Paolo Tomassetti (Aix-Marseille University and University of Milan); Joanna Unterschutz (University of Business Administration in Gdynia). Labour Market Law: Lilli Casano (University of Insubria); Silvia Spattini (ADAPT Senior Research Fellow). Social Security Law: Claudia Carchio (University of Bologna); Carmela Garofalo (University of Bari); Ana Teresa Ribeiro (Catholic University of Portugal – Porto); Alma Elena Rueda Rodriguez (National Autonomous University Of Mexico). Anti-discrimination Law and Human Rights: Erica Howard (Middlesex University) Anna Zilli (University of Udine). Labour Issues: Josua Grabener (Grenoble Institute of Political Studies); Habtamu Legas (Ethiopian Civil Service University); Francesco Seghezzi (ADAPT Senior Research Fellow). Language Editor Pietro Manzella (ADAPT Senior Research Fellow). Book Review Editors Peter Norlander (Loyola University Chicago). Scientific Committee of Reviewers Maurizio Del Conte (Bocconi University), Juan Raso Delgue (University of the Republic); Richard Hyman (LSE); Maarten Keune (University of Amsterdam); Felicity Lamm (Auckland University of Technology); Nicole Maggi-Germain (Pantheon-Sorbonne University); Merle Erikson (University of Tartu); John Opute (London South Bank University); Michael Quinlan (University of New South Wales); Jean Michel Servais (Honorary President of ISLLSS and Former Director of International Labour Office); Anil Verma (University of Toronto). E-Journal of International and Comparative LABOUR STUDIES Volume 11 No. 03/2022 @ 2022 ADAPT University Press Online Publication of the ADAPT Series Registration No. 1609, 11 November 2001, Court of Modena www.adaptbulletin.eu The articles and the documents published in the E-Journal of International and Comparative LABOUR STUDIES are not copyrighted. The only requirement to make use of them is to cite their source, which should contain the following wording: @2022 ADAPT University Press. Beyond Dualism: The Raise of Marginal Work in Italy Laura Maestripieri and Antonio Firinu * Abstract The ‘dualisation’ debate has given new emphasis to studies investigating labour market segmentation. However, we argue that the traditional dual divide between insider/outsider workers becomes inadequate when analysing the Italian labour market – which features a long series of reforms that segmented the labour market into several categories of workers. This paper aims to investigate the interrelation between the institutional dimensions of labour markets and their consequences on workers’ inclusion in social protection schemes, adopting an analytical framework to introduce an additional category of workers, i.e. marginal workers. Marginal workers are non-standard workers whose involuntary, intermittent and ambiguous character in the labour market prevents them from accessing social protection. Focusing on Italy, we provide empirical evidence, we put forward the following arguments a) marginal work is the result of a disjunction between social protection and non-standard work b) marginal work mainly concerns younger generations. One reason for this might be the systematic application of new dismissal rules to new entrants. Keywords: Dualisation; Labour market segmentation; Marginal work * Lara Maestripieri (corresponding author, email address: lara.maestripieri@uab.cat) is a distinguished researcher Ramon y Cajal at the Autonomous University of Barcelona (Spain). Antonio Firinu is an independent researcher based in Cagliari (Italy). E-Journal of International and Comparative LABOUR STUDIES Volume 11 No. 03/2022 @ 2022 ADAPT University Press - ISSN 2280-4056 BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 98 1. Introduction Labour market segmentation theories have recently been given new momentum following the success of the book “The Age of Dualisation” 1. However, recent studies have stressed how an analytical framework based on a dualistic view of labour markets – e.g., that related to the dualisation theory – falls short in analysing the complexity of the current labour markets. Research has demonstrated the simultaneous presence of different groups of outsiders with varying degrees of vulnerability. Today’s instability, fluidity and heterogeneity of careers diversify the extent to which people are exposed to labour vulnerability and call for a modulation of the dualisation theory2. However, in these studies, the link is missing as to how deregulation was enforced in labour markets and its outcomes in creating new segments of outsiders 3. In view of the above, the aim of this paper is to provide an analytical framework that conceives labour markets as a continuum, i.e., considering the shift from being an insider to holding multiple outsider positions. Building upon labour market segmentation and dualisation theories4, we conceptualise marginal workers as a distinct group among the outsiders. While outsiders are usually defined as “those with weak labour market integration involving non-standard employment forms”5, marginal workers are the most vulnerable ones among them. They are non-standard workers whose involuntary, intermittent and undefined character in the labour market limits access to social protection schemes. In this paper, Italy is considered as a relevant example of marginal work, due to the specific nature of deregulation in this country. First, the labour market reforms of the last 20 years have always followed the principle of 1 Emmenegger, P., Hausermann, S., Palier, B., & Seeleib-Kaiser, M. (2012). The age of dualization. The Changing Face of Inequality in Deindustrializing Societies. Oxford University press. 2 For example: Yoon & Chung, 2016; Pulignano e Doerflinger, 2018; Doerflinger et al., 2020; Seo, 2021. 3 Jessoula, M., Graziano, P. R., & Madama, I. (2010). ‘Selective Flexicurity’ in Segmented Labour Markets: The Case of Italian ‘Mid-Siders.’ Journal of Social Policy, 39(4), 561–583. Bureau, M.-C., & Dieuaide, P. (2018). Institutional change and transformations in labour and employment standards: An analysis of ‘grey zones.’ Transfer: European Review of Labour and Research, 24(3), 261–277. 4 E.g., Doeringer & Piore, 1971; Lindbeck & Snowden, 1988; Rueda, 2005; Emmenegger et al., 2012. 5 Fervers, L., & Schwander, H. (2015). Are outsiders equally out everywhere? The economic disadvantage of outsiders in cross-national perspective. European Journal of Industrial Relations, 21(4), 369–387. Citation at page 370. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 99 “deregulation at the margins”6. With this terminology, scholars indicate that dismissal rules have only been relaxed for new entrants to the labour market, skewing the chance of non-standard employment towards younger generations and leaving untouched the protection offered to those already employed when the reforms were enforced. Secondly, introducing more flexibility has not been adequately compensated by a transformation in the nature and entitlements of unemployment benefits 7. Since flexibilisation has always targeted new entrants, we argue that the marginal work created has affected younger generations more than older workers. The concept of marginal work draws from the previously recalled theories on labour market segmentation and dualisation. The idea of marginality was already present in Doeringer and Piore’s work about secondary labour markets (1971)8. Marginal work stresses that outsiders do not enjoy the same rights as insiders9 and is comparable with the similar “grey zone” concept10. Reflecting on Lindbeck and Snower (1988) 11, who theorised a different status for new entrants, we argue that there are different degrees of “outsiderness” and that younger generations tend to be more affected by extreme outsiderness (namely, what we call ‘marginal work’). This paper answers the following questions: what are the consequences of deregulation on the Italian labour market? Are these consequences unequally distributed across generations? In this sense, this research intends to deal with the following issues: a) to offer an analytical framework for understanding the rise of non- standard work in Italy, showing how a dualistic dichotomy between 6 Barbieri, P., Cutuli, G., Luijkx, R., Mari, G., & Scherer, S. (2019). Substitution, entrapment, and inefficiency? Cohort inequalities in a two-tier labour market. Socio- Economic Review, 17(2), 409–431. 7 Hacker, J. S. (2005). Policy drift: The Hidden Politics of US Welfare State Retrenchment. In W. Streeck & K. Thelen (Eds.), Beyond Continuity. Oxford University Press. 8 Doeringer, P. B., & Piore, M. J. (1971). Internal Labor Markets and Manpower Analysis. M.E. Sharpe. 9 Rubery, J., & Piasna, A. (2017). Labour market segmentation and deregulation of employment protection in the EU. In Myths of employment deregulation: How it neither creates jobs nor reduces labour market segmentation (p. 18). 10 Bureau, M.-C., & Dieuaide, P. (2018). Institutional change and transformations in labour and employment standards: An analysis of ‘grey zones.’ Transfer: European Review of Labour and Research, 24(3), 261–277. 11 Lindbeck, A., & Snower, D. J. (1988). The Insider-Outsider Theory of Employment and Unemployment. MIT Press. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 100 insiders/outsiders cannot grasp the complexity of the differences among workers; b) to analyse the long-term effect of policy implementation, to test the hypothesis that marginal work is a phenomenon that is unequally distributed across the generations. The ISTAT Italian Labour Force Survey (2009 – 2016) is the primary data source for the empirical analysis. The results of the investigation show that marginal work is a significant phenomenon in the Italian labour market and is unequally distributed across the generations, mainly affecting those under 36 years old. This paper is organised as follows. The first paragraph considers the debate on labour segmentation, starting from its first theorisation in the 1970s and connecting it to the more recent discussions on dualisation. The second paragraph analyses the deregulation process in Italy, highlighting how the deregulation process in Italy has been progressively put in place through various reforms – which started in 1997 and ended in 2014. The third paragraph defines the analytical framework for studying the insider/outsider divide in the Italian context. The fourth paragraph describes the data source and the variable operationalisation made to explore marginal work in the Italian labour market. The fifth paragraph illustrates the empirical evidence for our arguments, while the conclusions show the policy implications of the analysis of marginal work. 2. Dualism and Deregulation in Deindustrialised Labour Markets The first theorisation of dualism in labour markets dates back to Doeringer and Piore (1971) 12. They coined the term “secondary labour market” to refer to low-wage employment offered by small-size enterprises with informal and unstructured work opportunities. They distinguish it from the employment offered in the context of medium- and large-size enterprises (called the “primary labour market”), characterised by in-firm and regulative institutions protecting workers from instability. Workers with employment disadvantages, such as lower educational attainment and less job experience, tend to find work in the secondary labour market. Employment in the secondary labour market is characterised by instability because of frequent turnover, and work is not organised to provide continuous employment. In a later paper, Piore 12Doeringer, P. B., & Piore, M. J. (1971). Internal Labor Markets and Manpower Analysis. M.E. Sharpe. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 101 defines those workers in the secondary labour market as having a marginal labour force attachment or being affected by marginality 13. The insider/outsider theory proposed by Lindbeck and Snower led to the segmentation argument (1988)14. Their work emphasises conflicts between different types of workers caused by the core workers’ (insiders) capacity to protect their employment to the detriment of others (outsiders). They were the first to use the terms ‘insiders’ and ‘outsiders’: insiders are employees whose jobs are protected by labour turnover costs. In contrast, outsiders do not enjoy the same power and pay the price of labour adjustments. In other words, they explicitly link the bargaining power of insiders and their capacity to influence policymaking with outsider marginalisation. The two authors also argue that there are different degrees of “insiderness”: new entrants have a lower status as they have not yet acquired the same market powers as the more experienced insiders15. David Rueda took up the insider/outsider theory of employment and unemployment in 2005 in a completely different context to use it to explain partisanship in deindustrialised societies. He argued that differences between insiders and outsiders emerged during the industrial era when the economic growth of the 1960s and the union activism of the 1970s allowed some labour to be protected from unemployment (insiders) by reducing the employer’s ability to fire them. The labour supply shock of the 1980s, caused by the rising number of women entering the labour market, the increased international competition, and the flexibility required by changing consumer behaviours, incremented the number of people working outside the insider model. This led to an increasing number of part-time and temporary workers whom employers used to make labour adjustments that could not be done with insiders 16. Dualisation indicates this differential treatment of insiders and outsiders regarding access, rights and entitlements to protection from unemployment. Nevertheless, deindustrialisation did not trigger dualisation everywhere. It occurred in those countries that had protected economies with rigid employment protection in the industrial era, such as 13Piore, M. J. (n.d.). Labor Market Segmentation: To What Paradigm Does It Belong? 6. 14 Lindbeck, A., & Snower, D. J. (1988). The Insider-Outsider Theory of Employment and Unemployment. MIT Press. 15 Lindbeck, A., & Snower, D. J. (1988). The Insider-Outsider Theory of Employment and Unemployment. MIT Press. 16 Rueda, D. (2005). Insider–Outsider Politics in Industrialized Democracies: The Challenge to Social Democratic Parties. American Political Science Review, 99(1), 61–74. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 102 Italy. They proved highly vulnerable because they could not adapt smoothly to the economic shocks of the 1980s17. When they had to make their labour markets more flexible, they did so selectively by targeting those at the margins of the labour markets, such as new entrants, and protecting those already employed18. The main contribution of the more recent theory of dualisation is its attention to workers with non-standard contracts. Although Doeringer and Piore (1971) 19 had already emphasised the relation between non- standard jobs and secondary labour markets, Rueda (2005)20 explicitly included part-time and temporary workers in the category of outsiders. They are outsiders because of their lower entitlements to social rights, determined mainly by social security regulations still based on the standard employment relationship21. His argument is that non-standard workers, especially if involuntarily employed, suffer from gaps in the generosity and duration of social protection. Dualism in social protection 22 interrelates with the dual labour market structure23. The insiders enjoy high employment protection and job stability levels and are primarily concentrated in large firms and public employment (the primary labour market). The outsiders, mostly in non-standard or self-employed positions, in small- and medium-sized companies (secondary labour 17 Rueda, D., Wibbels, E., & Altamirano, M. (2015). The origins of dualism. In P. Beramendi, S. Häusermann, H. Kitschelt, & H. Kriesi (Eds.), The politics of advanced capitalism (pp. 1–40). Cambridge University Press New York. 18 Barbieri, P., Cutuli, G., Luijkx, R., Mari, G., & Scherer, S. (2019). Substitution, entrapment, and inefficiency? Cohort inequalities in a two-tier labour market. Socio- Economic Review, 17(2), 409–431. 19 Doeringer, P. B., & Piore, M. J. (1971). Internal Labor Markets and Manpower Analysis. M.E. Sharpe. 20 Rueda, D. (2005). Insider–Outsider Politics in Industrialized Democracies: The Challenge to Social Democratic Parties. American Political Science Review, 99(1), 61–74. 21 Fervers, L., & Schwander, H. (2015). Are outsiders equally out everywhere? The economic disadvantage of outsiders in cross-national perspective. European Journal of Industrial Relations, 21(4), 369–387. 22 Lindbeck, A., & Snower, D. J. (1988). The Insider-Outsider Theory of Employment and Unemployment. MIT Press. Rueda, D. (2005). Insider–Outsider Politics in Industrialized Democracies: The Challenge to Social Democratic Parties. American Political Science Review, 99(1), 61–74. 23 Doeringer, P. B., & Piore, M. J. (1971). Internal Labor Markets and Manpower Analysis. M.E. Sharpe. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 103 market), receive little protection from the regulations and low remuneration and are often exposed to job instability 24. In the last decade, there has been a multiplication of studies on labour market segmentation that called in question an analytical approach based on the two opposite poles (primary/secondary, insider/outsider) 25. Focusing on the Italian case, Jessoula et al. (2010) 26 stress the importance of the category of so-called mid-siders, which includes self-employed workers and employees of small businesses. They enjoy less protection than classic standard workers (large companies and public employment); moreover, the authors indicate among the outsiders also those who are excluded, even partially, from employment relationships, that is, the unemployed and discouraged. Yoon and Chung (2016) 27 found that in the UK, there is a third segment of the labour market whose workers will have insecure pension coverage despite their current permanent jobs. Doerflinger et al. (2020)28 evidenced the existence of five segments i n the European Labour Markets based on the insecurity divides that cut across traditional divisions between permanent and temporary workers. On the same line, Seo (2021) 29 found three different types of outsiders: typical outsiders – characterised by insecure employment, low income and lack of job prospects – dead-end insiders – those insiders whose job does not guarantee sufficient income and/or job prospects – and subjective outsiders – those workers who feel insecure about their jobs despite their permanent contract. The most interesting paper – at least in terms of the proximity to 24 Valadas, C. (2017). A Changing Labour Market under the Intensification of Dualization. The Experience of a Southern European Society. Social Policy & Administration, 51(2), 328–347. 25 See: Jessoula et al., 2010; Yoon & Chung, 2016; Bureau & Dieuaide, 2018; Pulignano e Doerflinger, 2018; Doerflinger et al., 2020; Seo, 2021. 26 Jessoula, M., Graziano, P. R., & Madama, I. (2010). ‘Selective Flexicurity’ in Segmented Labour Markets: The Case of Italian ‘Mid-Siders.’ Journal of Social Policy, 39(4), 561–583. 27 Yoon, Y., & Chung, H. (2016). New Forms of Dualization? Labour Market Segmentation Patterns in the UK from the Late 90s Until the Post-crisis in the Late 2000s. Social Indicators Research, 128(2), 609–631. 28 Doerflinger, N., Pulignano, V., & Lukac, M. (2020). The social configuration of labour market divides: An analysis of Germany, Belgium and Italy. European Journal of Industrial Relations, 26(2), 207–223. 29 Seo, H. (2021). ‘Dual’ labour market? Patterns of segmentation in European labour markets and the varieties of precariousness. Transfer: European Review of Labour and Research, 27(4), 485–503. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 104 our argument – is the one proposed by Bureau and Dieuaide (2018) 30. In their paper, they theorise the constitution of “grey zones”, in which laws are absent or weak due to the layering of several regulations. Although their interest is more addressed to the rising of new institutions, their approach is analytical, quite like the concept of marginal work, in which we propose to analyse the “grey zones” created by the Italian stratified deregulation process. In conclusion, the reviewed research has demonstrated that new categories of outsiders are rising31. We argue that it occurs because different groups of outsiders can access different levels of social protection: i. because they have different contracts and associated benefits; ii. because institutional regulations might change over time, altering the generosity of the benefits different generations of outsiders can access. The following section describes the process of deregulation that has occurred in Italy, showing how its outcome has been to differentiate the level of social protection reserved for different groups of outsiders across generations of workers. 3. The Nature of Italian Deregulation Italy is the ideal case for a study of labour market segmentation. It has a productive structure characterised by a limited primary labour market (owing to the prevalence of self-employment and firms with under 15 employees) and traditionally strong institutional barriers that divide insiders and outsiders32. The Italian case also stands out because the insider/outsider scenario is mainly based on an age/cohort divide. Barbieri et al. (2019) 33 argue that the deregulation process in Italy has been partial and targeted: the protections for insiders already in employment have remained unchanged through the various reforms. The deregulation has favoured an increase in non-standard employment, disproportionally 30 Bureau, M.-C., & Dieuaide, P. (2018). Institutional change and transformations in labour and employment standards: An analysis of ‘grey zones.’ Transfer: European Review of Labour and Research, 24(3), 261–277. 31 Studies like: Jessoula et al., 2010; Yoon & Chung, 2016; Bureau & Dieuaide, 2018; Pulignano e Doerflinger, 2018; Doerflinger et al., 2020; Seo, 2021. 32 Jessoula, M., Graziano, P. R., & Madama, I. (2010). ‘Selective Flexicurity’ in Segmented Labour Markets: The Case of Italian ‘Mid-Siders.’ Journal of Social Policy, 39(4), 561–583. 33 Barbieri, P., Cutuli, G., Luijkx, R., Mari, G., & Scherer, S. (2019). Substitution, entrapment, and inefficiency? Cohort inequalities in a two-tier labour market. Socio- Economic Review, 17(2), 409–431. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 105 targeting new labour market entrants, causing a generational effect on the younger cohorts34. The following historical reconstruction of the main deregulation reforms (from 1997 to 2014) aims to illustrate how this has occurred. The first deregulation law in Italy was the Treu Law, introduced on 24 June 1997 (Law 196/97). It aimed to increase labour market flexibility and reduce the unemployment rate during the centre-left Prodi government (1996-1998). Its most crucial intervention concerned legal provisions for the indirect acquisition of work through different non-standard forms of employment, namely agency work and para-subordinate work. In particular, it determined the abolition of the public monopoly on temporary agency work – introduced with Act No. 1369/1960 -and recognition of the legitimacy of private employment agencies. It liberalised the use of para-subordinate contracts such as “coordinated and continuous collaboration (co.co.co)” arrangements and introduced “co.co.pro” contracts, activated for a specific (and supposedly limited- duration) project. These contracts stood in-between dependent work and autonomous contracts, as they were suitable for hiring collaborators without granting them the rights of dependent workers (including unemployment benefits or vacations). These contracts were appetible for employers, as their cost was lowered due to the scarce contributions scheme granted to workers. Consequently, these workers faced precarious conditions since the beginning: the Italian labour protection did not cover them due to workers being de-jure self-employed but de facto employees35. In 2001, Law 368/2001 introduced temporary work in Italy as a reception of the EU directive 1999/70/CE. Temporary work was only available in case of technical, productive or organisational reasons, so the law required to state the reason behind the fixed term status of the contract. Later, the Biagi Law (no. 276, December 2003) further liberalised the employment regulation system. It introduced several non-standard contractual arrangements (such as on-call work, occasional work, and so on). However, adequate security schemes have not adequately compensated 34 Barbieri, P., Cutuli, G., Luijkx, R., Mari, G., & Scherer, S. (2019). Substitution, entrapment, and inefficiency? Cohort inequalities in a two-tier labour market. Socio- Economic Review, 17(2), 409–431. 35 Muehlberger, U., & Pasqua, S. (2009). Workers on the Border between Employment and Self-employment. Review of Social Economy, 67(2), 201–228. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 106 for the increased flexibility 36. Despite being formally covered in the event of unemployment, part-time, fixed-term and temporary agency workers faced insurmountable barriers to accessing ordinary unemployment benefits due to the strict eligibility requirements and the existence of minimum contribution thresholds. Parasubordinate workers were excluded from unemployment benefits37. This situation was not amended until 2012 with the Fornero reform (Law 92/2012), approved on 28 June 2012 during the Monti technical government (2011-2013) and under pressures caused by the sovereign debt crisis. The law relaxed the dismissal rules for standard employment for the first time since 1978 38. It also provided stricter regulation of those contractual arrangements that were particularly at risk of abuse, like coordinated and continuous project collaborations and bogus self- employment39. The Fornero reform introduced a separate unemployment benefit for non-standard workers, less generous but with fewer requirements. So, its impact remains ambiguous; it did intervene to reduce social security inequalities for non-standard work 40. The last reform introduced in Italy, the “Jobs Act” (Law 183/2014), addressed the two issues left unsolved after the Biagi law: contractual simplification and social protection for non-standard work. First, the law further eased the dismissal rules for permanent workers by introducing a new standard employment contract, the “increasing protection contract” (CTC), and cancelled the co.co.pro formula. The goal was to incentivise permanent CTC contracts instead, intending to simplify the contractual framework. However, the new dismissal rule was only introduced for the CTC contracts – following the pattern of deregulation for new entrants, 36 Sacchi S, Berton F and Richiardi M (2009) Flessibilità del lavoro e precarietà dei lavoratori in Italia: analisi empiriche e proposte di policy. Rivista Italiana di Politiche Pubbliche 2009(1): 33–70. 37 Madama, I., & Sacchi, S. (2007). Le tutele sociali degli occupati in nuove forme di lavoro. Un’analisi della prassi applicativa. Rivista Del Diritto Della Sicurezza Sociale, 2007(3). 38 Eichhorst, W., & Marx, P. (2020). How stable is labour market dualism? Reforms of employment protection in nine European countries. European Journal of Industrial Relations, 0(0). 39 Tiraboschi, M., & Rausei, P. (2012). Lavoro: Una riforma sbagliata. Ulteriori osservazioni sul DDL n. 5256/2012, Disposizioni in materia di riforma del mercato del lavoro in una prospettiva di crescita. Adapt University Press. 40 Sciarra, S. (2013). Flessibilità e politiche attive del lavoro. Note critiche sulla riforma Monti-Fornero. Giornale Di Diritto Del Lavoro e Di Relazioni Industriali, 139(2013), 471–488. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 107 like in the Treu and Biagi laws 41. The “Jobs Act” was preceded by the Poletti Decree (DL. 34/2014), which abolished any limitation for companies to use fixed-term contracts. The second aspect was the replacement of different social benefits with a single universal benefit 42. The NaSPI – the new acronym for unemployment benefit - increased accessibility to all types of non- standard workers. The benefit is centred on a reasonably inclusive contributory requirement: 13 weeks of contributions in the previous four years of work plus 30 days of contribution in the previous year. The benefit amount is 75% of the average monthly wage, up to a maximum of 1,300 euros, and the duration is half the number of weeks of contribution, up to a maximum of two years. Co.co.co workers do not enter the system of NaSPI, but, for the first time, it offered an unemployment benefits system for co.co.co contracts (DIS-COLL), with a maximum duration of six months. The succession of reforms that deregulated the Italian labour market can be fruitfully interpreted by applying the policy drift frame. As theorised by Hacker43, “policy drift refers to cases of institutional change that result not from ‘formal revision’, but from policies’ failure to adapt to shifts in their social or economic context”44. The inaction occurred between the Biagi law (2003) and the Fornero law (2012). During those nine years, the socio-economic conditions changed because of the increase in non-standard contracts determined by the Treu and Biagi laws. However, the unemployment benefits (left unchanged) were unfit to offer the same level of protection to all the non-standard workers. The Fornero reform (2012) and the Jobs Act (2014) issued new institutions to amend the situation ex-post. This ex- post correction was not enough to revert the erosion of social rights concentrated among the younger generations that previous laws had 41 Eichhorst, W., & Marx, P. (2020). How stable is labour market dualism? Reforms of employment protection in nine European countries. European Journal of Industrial Relations, 0(0). 42 Eichhorst, W., & Marx, P. (2020). How stable is labour market dualism? Reforms of employment protection in nine European countries. European Journal of Industrial Relations, 0(0). 43 Hacker, J. S. (2005). Policy drift: The Hidden Politics of US Welfare State Retrenchment. In W. Streeck & K. Thelen (Eds.), Beyond Continuity. Oxford University Press. 44 Béland, D., Rocco, P., & Waddan, A. (2016). Reassessing Policy Drift: Social Policy Change in the United States. Social Policy & Administration, 50(2), 201–218. Citation at page 201-202. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 108 triggered. It still enforced a separate system for co.co.co workers. It created a “grey zone”, as theorised by Bureau and Dieuaide (2018)45. 4. An Analytical Framework based on Marginal Work In the previous paragraph, we showed how the Italian deregulation process could be interpreted as an empirical example of the policy drift theorised by Hacker (2005) 46. While the general configuration of the labour market was changing due to the new rules regulating labour market entry, the social security system remained stable until the most recent reforms (2012 – 2014). This led to a progressive drifting of workers from the protection system, which would only be amended in recent years. We argue that those who drifted out of social protection because of the inaction of public regulation are the marginal workers. According to Kanbur (2007)47, marginality is a statement that needs to be defined in relation to some other group or category of society or an average standard. Standard workers enjoy the maximum benefits associated with participation in the labour market regarding access to welfare and economic reward thanks to their standard employment relationship. The centrality of standard work is also determined by the fact that it is still the most widespread form of employment (§ section 5), despite the steady rise in the percentage of non-standard work in recent decades48. Starting from this centre, the other groups (namely, the outsiders) occupy a weaker position within the labour market, characterised by higher social risks and lower social protection. In our theoretical framework, outsiders are composed of several groups. The first outsider group we find is voluntary non-standard workers, whom we assume to be less marginalised than involuntary non-standard workers, following Fervers and Schwander 45 Bureau, M.-C., & Dieuaide, P. (2018). Institutional change and transformations in labour and employment standards: An analysis of ‘grey zones.’ Transfer: European Review of Labour and Research, 24(3), 261–277. 46 Hacker, J. S. (2005). Policy drift: The Hidden Politics of US Welfare State Retrenchment. In W. Streeck & K. Thelen (Eds.), Beyond Continuity. Oxford University Press. 47 Kanbur, R. (2007). Conceptualizing Economic Marginalization. Key-notes for the Living at the Margins Conference. https://goo.gl/6eGupG. 48 Hipp, L., Bernhardt, J., & Allmendinger, J. (2015). Institutions and the prevalence of nonstandard employment. Socio-Economic Review, 13(2), 351–377. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 109 (2015)49. This group has chosen a non-standard working relationship, which could be for several different reasons (i.e., work-life balance, education, illness, etc.). On the other hand, the second group of outsiders, the involuntary non-standard workers, suffer from this condition and its disadvantage. Also, with a non-standard work contract and fewer entitlements to benefits compared to insiders, their marginalisation is higher than voluntary non-standard workers because they did not choose this condition50. However, they are usually given some degree of entitlement to social protection, according to the Jobs Act reform in 2014. Going further, we identify another group, namely marginal workers, who are the result of the drifting process that has characterised the deregulation reforms in Italy. Marginal workers are those non-standard workers whose integration in the labour market is so frail and intermittent that it enables them minimal access to social protection and unemployment benefits. They are localised at the margins of the labour markets, in a grey area between under-employment and self- employment51. They may be in such a grey legislative position that they have no entitlements. Lastly, in the area of exclusion from the labour market, we find unemployed and discouraged workers currently out of work. Depending on their previous contractual relationship, they may or may not be entitled to unemployment benefits. 49 Fervers, L., & Schwander, H. (2015). Are outsiders equally out everywhere? The economic disadvantage of outsiders in cross-national perspective. European Journal of Industrial Relations, 21(4), 369–387. 50 Rueda, D. (2005). Insider–Outsider Politics in Industrialized Democracies: The Challenge to Social Democratic Parties. American Political Science Review, 99(1), 61–74. Rueda, D., Wibbels, E., & Altamirano, M. (2015). The origins of dualism. In P. Beramendi, S. Häusermann, H. Kitschelt, & H. Kriesi (Eds.), The politics of advanced capitalism (pp. 1–40). Cambridge University Press New York. 51 Bureau, M.-C., & Dieuaide, P. (2018). Institutional change and transformations in labour and employment standards: An analysis of ‘grey zones.’ Transfer: European Review of Labour and Research, 24(3), 261–277. Seo, H. (2021). ‘Dual’ labour market? Patterns of segmentation in European labour markets and the varieties of precariousness. Transfer: European Review of Labour and Research, 27(4), 485–503. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 110 Figure No. 1. A model of labour market marginalization Source: Authors’ own elaboration The margin-centre metaphor is a powerful tool to identify those who progressively leak out of social protection in a situation where it is substantially more vulnerable than the rest of the non-standard workers. It is important to stress that not all non-standard workers are marginal workers (Allmendinger et al., 2013): although part-time and temporary workers enjoy fewer entitlements compared to standard workers 52, the social protection provided in the Jobs Act is now adequate to protect most non-standard workers from the risk of job loss, thus making a dualism that opposes standards/insiders and non-standards/outsiders less relevant. Therefore, as the following section will show, marginal work is a phenomenon that only involves specific types of non-standard workers. 5. Marginal Work Data Sources and Operationalisation 5.1 Marginal Work Operationalisation The effect of the combination of long-term deregulation policies “at the margins”, the new labour market and the social security framework, is to expose new entrants to the risk of becoming trapped in a situation of marginal work. To analyse this situation, this study focuses on three 52Hipp, L., Bernhardt, J., & Allmendinger, J. (2015). Institutions and the prevalence of nonstandard employment. Socio-Economic Review, 13(2), 351–377. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 111 situations that allow us to explore the heuristic potentiality of marginal work. These three situations have been theoretically derived by examining the contractual categories more exposed to social risk in light of the current labour market and social protection framework. Nevertheless, these categories align with the risk categories identified in the Eurofond reports (see Eurofond, 2017). The first is “marginal involuntary part-time” workers. Involuntary part- time work is a condition of underemployment and under-utilized labour; it is a phenomenon connected with business cycles, the structural characteristics of the labour market, and employers’ preferences 53. The more the non-standard contract is imposed by the counterpart and not voluntarily chosen, the more it increases the marginality of the person. Working fewer hours than desired implies reduced occupational and social insurance access. In addition, involuntary part-timers are more frequently found in low-skilled positions in the secondary sectors and are more exposed to temporary employment 54. The NaSPI introduced within the Jobs Act reform increased the accessibility of unemployment benefits for part-time workers. However, the amount and duration of their unemployment benefit are much lower than full-time workers. To fulfil the criterion of marginalisation in the event of job loss despite the unemployment benefit offered by the Jobs Act, the analysis focused on low-wage, involuntary part-time workers, defined as those who earn less than two-thirds of the median wage of part-timers in Italy (calculated in 2016). Workers in this category represent 2.55% of employed people in Italy (2016).55 A second position relates to those workers who are in an intermediate position between subordinate and independent work 56, so-called “para- 53 Insarauto, V. (2021). Women’s Vulnerability to the Economic Crisis through the Lens of Part-time Work in Spain. Work, Employment and Society, 35(4), 621–639. Velitiotis, M., Matzaganis, M., & Karakitios, A. (2015). Involuntary part-time employment: Perspectives from two European labour markets. (No. 15/02; IMPROVE Discussion Paper). https://goo.gl/KRM8gi 54 Velitiotis, M., Matzaganis, M., & Karakitios, A. (2015). Involuntary part-time employment: Perspectives from two European labour markets. (No. 15/02; IMPROVE Discussion Paper). https://goo.gl/KRM8gi 55 This percentage refers to all dependent workers who are employed for less than 30 hours per week and state that they are part-timers because they could not find a full-time job and not because they voluntarily chose a reduced working schedule. 56 Seo, H. (2021). ‘Dual’ labour market? Patterns of segmentation in European labour markets and the varieties of precariousness. Transfer: European Review of Labour and Research, 27(4), 485–503. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 112 subordinate workers”. Contract types such as coordinated and continuous arrangements (co.co.co formula) or project contracts (co.co.pro formula), voucher contracts, or occasional employment are peculiarities of Italian labour market regulations, producing the most vulnerable temporary workers. They do not have a subordinate work contract and thus are excluded from the rights associated with this (i.e., sick pay, holidays, a severance package). However, at the same time, they do not enjoy the opportunities offered by independent work, such as autonomy in providing services (in terms of hours and location) and the possibility of having more than one buyer57. Furthermore, no economic penalty is associated with dismissal, as these workers are not entitled to a severance package when the contract ends. Their condition of marginality lies in their grey position between temporary subordinate employment and self- employment. The Jobs Act banned project contracts but left the regulation for coordinated and continuous arrangements untouched. The law acknowledges the specific nature of the latter contract, and these workers have a dedicated unemployment benefit scheme (DIS-COLL) less generous than NaSPI. These workers represent 1.29% of employed people in Italy (2016). Finally, a third position under consideration is the “dependent freelance”, namely the freelancer in solo self-employment (i.e., without any employees58). Although independent work is likely to lead to high profits and autonomy for the worker, solo freelancers are frequently exposed to low income, insufficient coverage from social protection, and low pension contributions59. This is because freelance employment is usually considered an entrepreneurial activity, and, as such, the business risk is borne by the worker, who at the same time is not entitled to unemployment benefits as social protection does not cover loss of turnover60. Muehlberger, U., & Pasqua, S. (2009). Workers on the Border between Employment and Self-employment. Review of Social Economy, 67(2), 201–228. 57 Seo, H. (2021). ‘Dual’ labour market? Patterns of segmentation in European labour markets and the varieties of precariousness. Transfer: European Review of Labour and Research, 27(4), 485–503. 58 Hipp, L., Bernhardt, J., & Allmendinger, J. (2015). Institutions and the prevalence of nonstandard employment. Socio-Economic Review, 13(2), 351–377. 59 Hipp, L., Bernhardt, J., & Allmendinger, J. (2015). Institutions and the prevalence of nonstandard employment. Socio-Economic Review, 13(2), 351–377. 60 Eurofond. (2017). Aspect of non-standard employment in Europe. Publications Office of the European Union. https://goo.gl/xEM1R1. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 113 In this paper, we only consider those freelance workers with limitations on the autonomy of their activity to be marginal workers. Previous legislative interventions have tried to take account of the phenomenon of so-called “bogus self-employment” (Fornero Law, 92/2012) by focusing on the criterion of economic dependence on a single buyer, which is not possible to measure with current survey data. As a proxy, limitations on the autonomy of how and when freelance services are provided indicate that workers are potentially exposed to economic dependence, with limited control over the duration and stability of their working activity 61. Dependent freelancers represent 0.60% of employed people in Italy (2016). 5.2 Population and Data Sources This paper presents an analysis based on the Italian Labour Force Survey 2009-2016, 62 intending to evaluate trends in non-standard and marginal work in Italy in the eight years following the financial crisis. The population in the analysis is aged 15-64 years, and it is further subdivided into three main age classes: young - under 25 years old; young adults - between 25 and 36 years old; and adults - over 37 years old. These thresholds were selected based on the literature and the timing of the institutional reforms: 25 is considered the lower limit of the “prime age” for labour market integration63 as younger individuals are primarily inactive due to educational purposes, while 36 was the age in 2016 of those born in 1980, meaning they were 16 years old when the Treu Law (196/97) came into operation. In other words, the age of 36 is a symbolic threshold, marking the first generation to have spent their entire career in a deregulated labour market. 64 Age is an under-researched dimension of labour market stratification, yet it is highly relevant when it comes to the 61 Eurofond. (2017). Aspect of non-standard employment in Europe. Publications Office of the European Union. https://goo.gl/xEM1R1. 62 Given the small number of groups under investigation, the four 2016 trimesters were pooled. The analysis is based on the entire year of 2016. Graph 3 also provides a comparison up to 2009 in order to show the dynamic trend of marginal workers across the eight-year period. The choice of 2009 as the lower bound derives from the availability of in-depth information in the Italian Labour Force Survey, which is released specifically for research purposes. 63 Dieckhoff, M., & Steiber, N. (2012). Institutional reforms and age-graded labour market inequalities in Europe. International Journal of Comparative Sociology, 53(2), 97–119. 64 Law 30/2000 (Law Berlinguer) made 16 the compulsory school leaving age. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 114 effects of institutional reforms65. The analysis excludes migrants from non-OECD countries because of their segregation into low-skilled and poor manual occupations66. The empirical section will first present a descriptive analysis showing the distribution of the different types of work identified (standard, voluntary non-standard, involuntary non-standard and marginal work) and how they changed in the eight years. Finally, the analysis is completed by running a logistic regression to estimate the probability of being a marginal worker according to the year of birth and the year of the survey (2009 and 2016). The model is controlled by gender, educational level in four ISCED classes (less than ISCED 2, ISCED 3 (2 years), ISCED 3 (4 -5 years), ISCED 5-8), sector of activity in five classes (agriculture, manufacturing, construction, advanced business services and other services), and geographical distribution based on five classes (North-West, North-East, Centre, South and the Islands). 6. Empirical Evidence about Marginal Work Marginal work affects 4.43% of the total labour force in Italy, an estimated 900,000 workers (Table 2). Although the standard employment relationship still represents the majority of employment (54.9%), there is quite a consistent percentage of non-standard employment: part-time, temporary contracts and self-employment account for approximately 40% of employment in Italy, of which almost one-third is involuntary (14.5% of total). However, the numbers concerning involuntariness can be misleading, as the Italian Labour Force Survey does not ask self-employed workers if they have chosen or were forced to opt for an autonomous job. In line with what is supposed by the segmentation theory, marginal workers feel insecure about their job. Among persons who believe they are likely to lose their job in the next six months, the proportion of marginal involuntary part-timers (26.8%) and para-subordinate workers (34.6%) is relatively high, especially compared to dependent freelancers (11.4%). Contrary to what Doeringer and Piore (1971) supposed67, marginal work is not only associated with jobs in the secondary sectors or 65 Dieckhoff, M., & Steiber, N. (2012). Institutional reforms and age-graded labour market inequalities in Europe. International Journal of Comparative Sociology, 53(2), 97–119. 66 Fellini, I. (2015). Una «via bassa» alla decrescita dell’occupazione: Il mercato del lavoro italiano tra crisi e debolezze strutturali. Stato e Mercato, 105, 469–508. 67 Doeringer, P. B., & Piore, M. J. (1971). Internal Labor Markets and Manpower Analysis. M.E. Sharpe. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 115 lower hierarchical positions (except for involuntary part-time work). In the case of para-subordinate (34.9%) and dependent freelancers (41.1%), more than one-third of the workers have at least tertiary education (see table A3 in the annexe). Para-subordinate workers and dependent freelancers make up higher percentages in advanced business services. In contrast, involuntary part-time workers are strongly associated with traditional services such as real estate activities, retail, accommodation and restaurants (see annexe, Table A2). It is also interesting to note that marginal employment is almost non-existent in the manufacturing and construction sectors, except for dependent freelance contracts, which probably disguise bogus self-employment in construction firms, which in theory is forbidden by law (see Fornero Law 2012). The different distribution of the types of marginal employment is also evidenced by the professional positions occupied by different workers. Part-timers are more likely to be employed as skilled workers in services and as unskilled workers. Freelancers have a dual distribution: one is concentrated on unskilled workers; the other is executives, intellectuals and office workers. Para-subordinate workers, by contrast, are more homogenous, as they are primarily active in service positions (skilled or highly skilled). Upon analysing the evolution of marginal work over time (Table 2), it can be observed how the percentage of marginal work remains stable, at approximately 4% of total employment, even though legislators have tried to reduce the phenomenon in recent years, for example by discouraging the use of bogus self-employment (Law Fornero, 2012) and abolishing co.co.pro contracts (Jobs Act, 2014). The result has been a progressive shift of marginal workers from dependent freelance and para-subordinate workers to involuntary part-time workers. In contrast, the total number of marginal workers is stable at around 900,000. In 2016, involuntary part- time workers accounted for two-thirds of the total marginal workers, increasing from only one-third in 2009. Given the downgrading trend suffered by the Italian labour market after the financial crisis68, it can also be reasonably argued that this progressive movement from dependent freelance and para-subordinate work to involuntary part-time work is not only due to the effect of the labour reforms but is also associated with a contraction in the labour demand for high-skilled positions and advanced business services following the years of crisis, which has reduced the number of para-subordinate and freelance workers in recent years. 68 Fellini, I. (2015). Una «via bassa» alla decrescita dell’occupazione: Il mercato del lavoro italiano tra crisi e debolezze strutturali. Stato e Mercato, 105, 469–508. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 116 Table No. 2. Distribution of Marginal Work: Percentages and Absolute Values Dependent Para- Involuntary % of Total Absolute Freelance subordinate Part-Time Employment Values Workers (estimation) Year 2009 24.08 42.99 32.93 4.10 856,881 Year 2010 24.94 40.99 34.07 4.43 911,075 Year 2011 18.53 45.05 36.43 4.20 858,908 Year 2012 20.42 41.48 38.11 4.64 941,133 Year 2013 18.13 39.76 42.11 4.54 895,719 Year 2014 17.95 36.9 45.15 4.66 913,436 Year 2015 14.23 32.41 53.36 4.68 922,038 Year 2016 13.38 28.96 57.65 4.43 883,546 Source: Authors’ own elaboration from the Italian Labour Force Survey (2009 – 2016) Figure No. 3. Distribution of work by model of marginalization: percentages (2009-2016)69 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Under25 (2009) 9,9 3,7 8,2 2,0 7,5 10,0 58,8 Under25 (2010) 8,6 3,4 8,3 2,2 8,3 10,1 59,1 Under25 (2011) 7,6 3,0 8,4 2,1 7,9 10,3 60,6 Under25 (2012) 6,4 3,2 8,7 2,3 10,2 10,5 58,7 Under25 (2013) 5,6 2,8 7,8 2,2 11,0 10,1 60,7 Under25 (2014) 4,8 2,5 8,0 2,1 11,9 10,0 60,7 Under25 (2015) 4,9 2,5 7,7 2,3 11,1 10,2 61,3 Under25 (2016) 5,6 2,5 8,4 2,4 10,5 9,7 60,9 Standard employment 25-36 (2009) 40,1 17,0 9,9 3,6 7,2 9,4 12,7 Voluntary Non-standard 25-36 (2010) 38,6 16,9 10,0 3,7 8,1 9,8 12,9 25-36 (2011) 37,8 16,3 11,4 3,7 7,8 10,1 12,9 Involuntary Non-standard 25-36 (2012) 35,8 15,3 12,4 4,1 10,6 9,7 12,2 Marginal izati on area 25-36 (2013) 33,5 14,7 12,7 3,7 12,0 10,4 13,0 Unemployed 25-36 (2014) 32,5 13,9 13,2 4,0 12,3 11,5 12,6 Discour aged 25-36 (2015) 32,0 14,1 13,3 4,1 12,4 11,7 12,4 Inactive 25-36 (2016) 32,3 14,3 14,0 3,9 12,0 11,4 12,1 Over 37 (2009) 36,3 18,6 4,6 2,0 3,0 5,9 29,6 Over 37 (2010) 35,9 18,6 4,8 2,2 3,3 6,3 29,0 Over 37 (2011) 35,8 18,7 5,4 2,0 3,3 6,4 28,5 Over 37 (2012) 36,0 18,4 5,8 2,2 4,5 7,0 26,0 Over 37 (2013) 36,1 17,9 6,2 2,2 5,2 7,1 25,3 Over 37 (2014) 36,5 17,8 6,5 2,3 5,5 8,0 23,4 Over 37 (2015) 36,9 18,2 6,5 2,3 5,3 8,4 22,4 Over 37 (2016) 37,6 18,6 6,6 2,2 5,2 8,1 21,7 Source: Authors’ own elaboration from the Italian Labour Force Survey 69 Discouraged workers are those who would like to work but are not currently actively looking for a job. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 117 Looking at Figure 3, the marginal work dynamic is influenced by the generation of workers. For older workers, the eight years are characterised by a slight growth in the rate of standard employment, to approximately 37% of the population. In contrast, the inactivity rate shows a contraction (-36.7%), primarily due to a reduction in the number of retired persons (from 10.2% of people over 37 in 2009 to 5.5% in 2016 - see Table A3 in the Annex). It can be assumed that these trends among older workers are the result of the pension reforms of the last decade, which aimed to increase the permanence of workers within the labour force (Laws 122/2010 and 148/2011, Sacconi Laws; Decree Law 201/2011, Fornero Law). However, instead of promoting employment (+3.7% for standard employment), the pension reforms have increased involuntary non- standard (+30.7%) and marginal work (+7.9%) and exclusion from the labour market (+43.3% for unemployment and +26.8% for discouragement). Nevertheless, the proportion of standard work still makes up the majority of workers over 37. For the youngest workers, the under-25s, the trend reveals a significant contraction in standard employment. Although this is most significant for the under-25s (-78%) compared to the other age groups, the fact that under-25s are, for the most part, still involved in education (approximately 60% of the total population of that age group, as evidenced in Table A3) makes the phenomenon less worrying in terms of its effect on the labour market. The number of under-25s outside the labour market has been stable, growing only slightly (+3.5%). In the most recent years, unemployment (+28.3%) and marginal work (+16.5%) grew amongst this group, but the percentage of total employment accounted for by marginal work (2.4%) is still less than in the prime age group (3.9%). Workers between 25 and 36 years old are especially exposed to marginal work. For this group, the period saw a loss of approximately one-quarter of the total standard employment, decreasing from 40.1% of the total population in 2009 to 32.2% in 2016. At the same time, exclusion from the labour market increased, with 65.7% more unemployed and 21.7% more discouraged in 2016 compared to 2009. The process of drifting is evident when looking at the differences between the last six years: standard (-19.6%) and voluntary non-standard positions (-16.2%) were lost among this age group. The financial crisis undoubtedly magnified this process but affected this age group more than older workers. In the same years, a corresponding growth was measured in involuntary non-standard (+41.4%) and marginal work (+9.9%); in 2016, the two accounted for approximately 18% of the population of 25- to 36-year-olds, an increase of 4.5% from 2009. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 118 The results of this descriptive analysis are further confirmed by analysing the effect of the year of birth on the probability of being marginally employed via logistic regression analysis. The younger generations always experience a higher likelihood of marginal employment, which is greater than previous generations when they were the same age. Looking at Figure 4, the two lines for 2009 and 2016 show the movement towards marginal work in 2016; however, the magnitude of the drift is not equa lly distributed across the age brackets and is more pronounced for the youngest generations (the graph only takes into account the generations of workers who were of working age in the two reference years, i.e., those born between 1953 and 1993). This is confirmed when controlling for gender, education, sector of activity and geographical residence (see Table A3 in the annexe). Notably, the gaps between 2009 and 2016 close approximately at the age of 36, as already mentioned, the first generation to have lived their entire working life in a deregulated labour market. For workers aged 37 and older, there is no empirical evidence of a difference between 2009 and 2016. Although the highlighted growth could be explained as an effect of the financial crisis, it is nevertheless evident that older workers were not affected by the crisis in the same way as the young and that deregulation led to a deterioration in labour market conditions borne mainly through younger workers. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 119 Figure No. 4. Predicted probabilities of being a marginal worker by age, 2009 and 2016 70 Source: Authors’ own elaboration from the Italian Labour Force Survey In conclusion, the empirical evidence presented in this section shows that the likelihood of being a marginal worker is more significant for those born after 1980, who have paid the cost of deregulation at the margins. In addition, marginal workers do not easily fit the profile of workers with an employment disadvantage: many are high-skilled, working in an advanced business sector and white-collar jobs. One might argue that marginal work results from individual inadequacy or a worker’s lower capacity. However, this interpretation does not fit the profile of Italian marginal workers with human capital to offer in the labour market and are still young enough to be more productive than the average adult worker. Para-subordinate workers and dependent freelancers – two of the three categories of 70The graph shows the predicted probability by age and year, based on the model shown in Table A4 in the Annex, which controls for gender, education, sector of activity and geographical location. Although the model does not include age but year of birth, the authors consider it legitimate to use age in the graph as the two variables are perfectly collinear (correlation values -0.96): age was selected as it is more informative for the argument of this paper. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 120 marginal workers identified in this paper – are strongly present among the highly skilled positions in the labour market, in contrast with the theory of labour market segmentation. Work experience does not seem a possible explanation since marginal work is also a significant phenomenon among the over-30s, not just the younger new entrants. It might be more realistic to argue that marginal work is a legacy of how deregulation has been implemented in Italy because of the disjunction between the social protection schemes and the changed economic conditions for workers, as predicted by the policy drift theory 71. 7. Discussion and Conclusions This paper makes two main contributions to labour market segmentation theory and the more recent dualisation theory. First, following recent research72, our paper evidenced the limitations of a dualist vision of the opposed insider/outsider poles, by highlighting that at least in Italy the outsiders consist of different groups. One of them, namely the marginal workers, consists of those who have drifted out of the social protection system73 owing to inaction on the part of the government in making the necessary updates to benefits in light of the progressive introduction of more flexible contracts. Our empirical evidence confirms previous research74, which theorised the importance of stratified deregulation processes in determining areas of work characterised by weak or absent social protection. Second, we have given empirical evidence for our hypothesis that younger generations are more exposed to marginal work due to the specific nature of deregulation in Italy 75. Through an empirical analysis of the Italian 71 Hacker, J. S. (2005). Policy drift: The Hidden Politics of US Welfare State Retrenchment. In W. Streeck & K. Thelen (Eds.), Beyond Continuity. Oxford University Press. 72 See: Jessoula et al., 2010; Yoon & Chung, 2016; Pulignano & Doerflinger, 2018; Doerflinger et al., 2020; Seo, 2021. 73 Hacker, J. S. (2005). Policy drift: The Hidden Politics of US Welfare State Retrenchment. In W. Streeck & K. Thelen (Eds.), Beyond Continuity. Oxford University Press. 74 Bureau, M.-C., & Dieuaide, P. (2018). Institutional change and transformations in labour and employment standards: An analysis of ‘grey zones.’ Transfer: European Review of Labour and Research, 24(3), 261–277. 75 Barbieri, P., Cutuli, G., Luijkx, R., Mari, G., & Scherer, S. (2019). Substitution, entrapment, and inefficiency? Cohort inequalities in a two-tier labour market. Socio- Economic Review, 17(2), 409–431. www.adapt.it LAURA MAESTRIPIERI AND ANTONIO FIRINU 121 Labour Force Survey 2009-2016, this paper has demonstrated that a considerable proportion of non-standard workers drifted from standard employment in the eight years, a phenomenon particularly affecting the population of young adults (25-36 years old). This research confirms previous empirical evidence in literature. The role of institutions in generating multiple categories of outsiders was already put in evidence by Jessoula et al., in 2010 76, by analysing those categories of workers which stood in-between insiders and outsiders in the Italian system. However, differently from them and similar to Yoon and Chung (2016)77, our analysis showed how the process of implementing deregulation in Italy has also determined the progressive drifting of specific categories of workers outside the social protection systems, with a clear generational divided determined by “deregulation at the margins” approach78. Secondly, compared to studies like Doerflinger et al. (2020)79 and Seo (2021) 80, we take a step further in including more workers among outsiders. Our analysis does not stem from a subjective perception of the worker regarding their job insecurity as in the cited research. We consider that being involuntarily employed in part-time work is a determinant factor for being a marginal worker together with dependent self- employment81, as it predicts low-income revenue and scarce job prospects 76 Jessoula, M., Graziano, P. R., & Madama, I. (2010). ‘Selective Flexicurity’ in Segmented Labour Markets: The Case of Italian ‘Mid-Siders.’ Journal of Social Policy, 39(4), 561–583. 77 Yoon, Y., & Chung, H. (2016). New Forms of Dualization? Labour Market Segmentation Patterns in the UK from the Late 90s Until the Post-crisis in the Late 2000s. Social Indicators Research, 128(2), 609–631. 78 Barbieri, P., Cutuli, G., Luijkx, R., Mari, G., & Scherer, S. (2019). Substitution, entrapment, and inefficiency? Cohort inequalities in a two-tier labour market. Socio- Economic Review, 17(2), 409–431. 79 Doerflinger, N., Pulignano, V., & Lukac, M. (2020). The social configuration of labour market divides: An analysis of Germany, Belgium and Italy. European Journal of Industrial Relations, 26(2), 207–223. 80 Seo, H. (2021). ‘Dual’ labour market? Patterns of segmentation in European labour markets and the varieties of precariousness. Transfer: European Review of Labour and Research, 27(4), 485–503. 81 Muehlberger, U., & Pasqua, S. (2009). Workers on the Border between Employment and Self-employment. Review of Social Economy, 67(2), 201–228. Hipp, L., Bernhardt, J., & Allmendinger, J. (2015). Institutions and the prevalence of nonstandard employment. Socio-Economic Review, 13(2), 351–377. Seo, H. (2021). ‘Dual’ labour market? Patterns of segmentation in European labour markets and the varieties of precariousness. Transfer: European Review of Labour and Research, 27(4), 485–503. @ 2022 ADAPT University Press BEYOND DUALISM: THE RAISE OF MARGINAL WORK IN I TALY 122 – already highlighted by Seo (2021)82 as critical elements for segmentation. Although being an involuntary non-standard employed worker is a subjective condition as well, plenty of empirical evidence has demonstrated that such a condition exposes to a problematic integration into the social protection system, determined by the reduced number of working hours, which impacts their social contributions83. This analysis has important policy implications. The problem of younger generations depends on their unequal exposure to deregulation, which is different among the generations coming after the first reforms compared to those already employed. The insider/outsider divide thesis supporters have encouraged the relaxation of the norms against dismissal in standa rd employment relationships to favour transitions between insiders and outsiders. However, this strategy only spreads precariousness rather than dealing with the marginality suffered by outsiders 84. Along with most Italian non-standard workers, the condition of marginal workers is involuntary: they accept these types of jobs because jobs with good working conditions (full-time and dependent) are not available. A universal benefit like a minimum income should give these workers sufficient bargaining capacity to refuse underpaid and precarious jobs. If this policy looks unrealistic in the current economic outlook, at least extending non-standard workers’ access and entitlements to active labour market policies should sufficiently empower them to find a job with better conditions. In conclusion, this analysis confirms the consolidation of multiple segmentations within the labour market in Italy, in which the traditional divide between insiders and outsiders has yielded more fragmentation than a simple dualism and a situation of marginality mainly concentrated among the younger generations. 82 Seo, H. (2021). ‘Dual’ labour market? Patterns of segmentation in European labour markets and the varieties of precariousness. Transfer: European Review of Labour and Research, 27(4), 485–503. 83 Insarauto, V. (2021). Women’s Vulnerability to the Economic Crisis through the Lens of Part-time Work in Spain. Work, Employment and Society, 35(4), 621–639. 84 Rubery, J., & Piasna, A. (2017). Labour market segmentation and deregulation of employment protection in the EU. In Myths of employment deregulation: How it neither creates jobs nor reduces labour market segmentation (p. 18). www.adapt.it ADAPT International Network ADAPT is a non-profit organisation founded in 2000 by Prof. Marco Biagi with the aim of promoting studies and research in the field of labour law and industrial relations from an international and comparative perspective. Our purpose is to encourage and implement a new approach to academic research, by establishing ongoing relationships with other universities and advanced studies institutes, and promoting academic and scientific exchange programmes with enterprises, institutions, foundations and associations. In collaboration with the Centre for International and Comparative Studies on Law, Economics, Environment and Work, (DEAL) the Marco Biagi Department of Economics, University of Modena and Reggio Emilia, ADAPT set up the International School of Higher Education in Labour and Industrial Relations, a centre of excellence which is accredited at an international level for research, study and postgraduate programmes in the area of industrial and labour relations. Further information at www.adapt.it. For more information about the E-journal and to submit a paper, please send a mail to LS@adapt.it.