Working in hospitality and catering in Greece and the UK: Do trade union membership and collective bargaining still matter?

Existing literature suggests that terms and conditions of employment are universally poor in hospitality and catering. Based on fieldwork comprising more than 70 semi-structured in-depth interviews conducted in the UK and Greece in 2018 and 2019, this article maps the field in the two countries and discusses the key parameters that structure working conditions, pay determination, and the impact of trade unionism and absence thereof. Poor working conditions and precariousness are prevalent in both contexts as expected, but surprisingly wages and other terms and conditions at least for a significant section of the hospitality workforce are comparatively better in Greece. An explanation for these findings is suggested with reference to the tradition of collective bargaining in hotels in Greece which partially survived the 2008 crisis and the legal reforms that followed. Trade unionism, despite its weakening and discrediting, has left its mark as a tradition contributing to the emergence of formal labour-based initiatives and more broadly to maintaining an openness to collectivism and collective action, more so than in the UK context. This combination of the distinct institutional and sectoral features of the Greek case and worker agency has served to ameliorate the context and content of employment for substantial sections of the hospitality workforce.


Introduction
My salary is fixed. In 2017, they gave us a pay rise and it's £19,556 now … which is not too much. We have some benefits but it's not very much. When I started in this company, they gave us, for instance, leave and extra weekends every year and they score out of that, we don't have that anymore since three years ago that we are part of a new company (Male, 55, non-British, night receptionist).
Hours were fully paid according to the sector agreement. I used to have a stable working pattern without much fluctuation. My wage could reach €1500 per month which was double the then minimum wage rate (Female,53,Greek,cleaner). These quotes speak of quite different working conditions, pay levels, and benefits. This seems to contrast with the common perception in the literature that hospitality work is invariably low paid, unstable, low-skilled, seasonal, and precarious across different geographical and jurisdictional contexts mainly because structural factors such as competitive pressures and 'price wars' among hospitality businesses, especially large ones, prevent any substantial improvement in working conditions (Baum, 2015;Kusluvan, 2003;Lloyd and Payne, 2021;Vanselow et al., 2010). Seeking to interrogate conventional opinion about common bad employment trends in hospitality, we researched the sector in the UK and Greece. The main objective was to understand the ways in which the different national institutional contexts mediate structural pressures on hospitality workers' terms and conditions and affect union responses. With that aim in mind, we examined a range of workplaces (large and small, seasonal and non-seasonal), occupational groups (housekeeping, chefs, restaurant and bar workers, and receptionists), and contractual statuses (permanent-temporary, seasonal, outsourced, etc).
Existing scholarship has largely focused on working conditions and collective action in hospitality and catering within the confines of a particular country and has rarely compared countries with different socio-economic systems, mainly overlooking the possibility of diverse outcomes in terms of employment quality and opportunities-chances for unionization across different countries. Comparative research on low-paid work and working conditions in hospitality has indicated the relevance of national institutions (Gautié and Schmitt 2010;Jordhus Lier and Underthun 2015;Lloyd and Payne, 2021) but has remained rare. To address this gap, which leads to a neglect of questions of context and power, we investigate and analyze both the similarities among hospitality and catering workers' experiences in these two different contexts as well as where those experiences differ and why. In doing so, we assess the impact of the institutional order on terms and conditions of employment and search for the structural and agentic factors that drive convergence or divergence. Current studies examine the impact of national institutions mainly by looking at union density and collective bargaining systems, but there is much less focus on the importance and impact of different union cultures on the ability of workers to mobilize and respond to employers' strategies even in unfavourable institutional contexts.
A comparative study of employment and industrial relations in different European countries can contribute to an understanding of the growth of poor working conditions (Baum, 2015) already evident in hospitality. With growing precariousness and eroding labour rights in the hospitality industry, identifying the possibilities for collectivism and solidarity in different industrial relations traditions and regulatory frameworks can both enrich employment relations theory and inform employment relations practice for progressive change in a low-paying sector of the economy. Through the investigation of two countries/contexts which have been under great strain in recent years, we demonstrate that historical and institutional factors influence the capacity for worker agency and consequently the likelihood for improved working conditions. Thus, even in the most adverse conditions, if appropriate historical and institutional elements are utilized, there exists the potential for power mobilization and scope for winning better terms of employment.

Literature review
The hospitality and tourism sector has been characterized by poor working conditions such as job insecurity, work stress, low well-being, lack of work-life balance and emotional exhaustion (Gautié and Schmitt 2010). The main drivers behind these converging trends include growing political instability, globalization, heightened competitive pressures, neoliberal policies, and the dominance of large corporations (Baum, 2015;Bianchi, 2011;Lopez et al., 2019). By making the global environment more volatile and unpredictable, these drivers force hospitality businesses across different countries to compete on a larger scale, respond to sudden demand fluctuations and seek profitable/ cost-efficient investments through constant restructuring and new organizational models (Eurofound, 2012;Vanselow et al., 2010). These trends have exacerbated pre-existing cost-reduction and flexibilization strategies, associated with the seasonal and labourintensive nature of the industry and the availability of a large number of migrant workers and vulnerable groups (women and young) who have grown in recent years and are willing to accept low-paid positions. In addition, neoliberal globalization has dismantled barriers to cross-border capital and labour mobility and eroded labour regulations, leading to leverage of capital power and weakening of labour's ability to resist (Bianchi, 2011:9). Union organizing and collective action have become even harder across different countries due to the individualization of the employment relationship, the increasing complexity of corporate structures (outsourcing, subcontracting), and the limited collective bargaining coverage across EU countries (Lopez et al., 2019).
It has been recognized though that little attention has been paid to how institutions and worker mobilization can play a positive role in alleviating working conditions in the lower end of the service sector labour market (Gray, 2004). According to this argument, unions and institutions cannot transform industry-specific conditions such as low-skilled and routinized work but can structure the hospitality labour market in ways that pay and conditions of employment substantially improve (Alberti, 2016;Bergence et al., 2015). Following Dunlop (1958: 94), this article highlights that while the technical and market contexts of an industry may be similar across countries, the 'power context' and the 'status of the actors' -'workers and their organizations, managerial hierarchies, and governmental agencies'vary significantly. This power context has recently been under significant pressure because of the enactment of labour market and welfare reforms that profoundly erode the ability of institutions and unions to improve working conditions (Ioannou 2021). However, we argue that there is still little attention paid by current research to whether and, if so, how unions can use past legacies to mobilize workers and collectively act even in adverse institutional settings. More emphasis needs to be placed on the intersection between past legacies/union power resources and sectoral dynamics that condition workers' agentic capacity to cement worker solidarity and resist work degradation. To our knowledge, however, existing studies on employment relations and trade unions in the hospitality industry focus on the micro-level including individual workplaces or regions (Wills, 2005;Jordhus Lier and Underthun 2015;Bergence et al., 2015) but without considering the wider institutional-political and economic context within which hospitality work is located.

The comparative rationale
Greece and the UK have been categorized into different systems of industrial relations (IR), with the former typically associated with Mediterranean capitalism and the latter as an example of a market-based economy (Amable, 2003;Georgen et al., 2012). The Greek IR system has been synonymous with comparatively low instance of flexible work, national-level collective bargaining, sectoral-occupational agreements, and high levels of coverage (Karamessini, 2015). However, the Greek IR system was seriously dismantled after 2010 due to the implementation of austerity and 'internal devaluation' policies that aimed at increasing economic competitiveness and avoiding State bankruptcy. This exceptionally drastic and forced change introduced new features into the system such as the decentralization of collective bargaining, the suspension of the extension principle, and the right of unorganized workers to conclude company agreements (Kouzis and Kapsalis, 2020).
These changes not only marginalized the role of unions in Greek IR but also repositioned Greece in the European IR landscape. Eurofound (2018a) considers Greece as the only country changing cluster between 2008-2017 moving from the State Centred Model of high collective bargaining coverage with State intervention to a voluntaristic associational governance model characterized by low collective agreement coverage and weaker voluntaristic associational governance. The UK on the other hand has been clustered into the market-oriented group of countries owing to limited State intervention, low collective bargaining coverage and decentralized (and uncoordinated) wage-setting systems (Eurofound, 2018a). In hospitality, after the decentralization and weakening of collective bargaining, Greece has moved closer to the UK since most employees are not covered by collective agreements (Eurofound, 2018b). Another similarity is that unions are not consulted on sector-related issues while employers' associations seem to be more involved. Greek employers have reported higher levels of consultation and influence compared with their UK counterparts (Eurofound, 2018b: 34).
Nonetheless, there are also enduring differences between the two contexts that may hinder convergence. Among these features are the role and ideological orientation of unions both in hospitality and beyond. Although most Greek unions have adopted consensus-based approaches, industrial conflict is still higher than in the UK and for historical reasons (district political legacy-history) the labour movement is more politicized, and militant-class oriented tendencies coexist with more moderate ones, which in theory allows some space for more agentic action even in unfavourable employment and regulatory conditions (Kritsantonis, 1998;Papadopoulos and Lyddon, 2020). UK hospitality on the other hand has historically been rather averse to union organizing with very low union rates, failed attempts to organize workplaces and absence of sectoral collective agreements (Lucas, 2009;Wood, 2020). We aim to examine whether and to what extent the Greek industrial relations context exercises some influence in favour of imposing constraints on the degree of precariousness that hospitality workers experience. We thus theoretically and empirically engage with Hyman's (2018) question regarding the dialectic between structure and agency; can agency resist adverse institutional settings and fight precariousness in low-paying sectors of the economy?
The assumption about the significance of national institutions for explaining differences in job quality in low-paying sectors (Gray, 2004) needs to be empirically investigated by also elaborating on sectoral dynamics. In a recent study examining precarious work in three low-skilled sectors across different countries, Keune and Pedaci (2020) concluded that sectors show similarities across countries since national institutional differences are less important than similar sectoral paths. The authors highlight the importance of power resources in partially ameliorating precariousness with an emphasis on union strength. Our study complements their analysis and fills in a critical gap concerning the significance of understanding specific features of the sector in each country that might enable or inhibit the agency of workers to develop. We argue that although path dependency can explain cross-national similarities in sectors with increasing weight and significance in the national economy, some national features can have a stronger impact than others. The comparison of the two countries explores the role of regional-sectoral diversity and responds to the call by Carre and Tilly (2012) to 'examine a single sector across countries' which has been seriously neglected in low-pay service work.
Contrary to the dramatic post-2008 decline in the Greek economy, Greek hospitality and catering reported significant growth especially since 2013 mainly because of substantial tourist growth. The sector, despite the recent increase of four-and five-star hotels (some of which are large, branched, and/or franchised), remains dominated by small and medium enterprises and decentralized ownership models with around 80% of firms having fewer than 50 employees (Insete, 2020). In the UK, there has been a significant increase in larger establishments in recent years with brands and franchises occupying a greater share of the market than in Greece, while many microenterprises are part of big international chains (Ioannou and Dukes, 2021). The UK hotel sector has a high concentration with around 15 brands controlling 40% of the total available rooms, while an asset-light model combining different adjustments between ownership, lending and management is dominant (Lopez-Andreu et al., 2019). Owing to different levels of capitalist growth and position in the division of labour there are also differences between the position of the sector in the two countries. In Greece, hospitality and tourism occupy a central position in the economy contributing around 10% of GDP and 20% of employment but remain largely seasonal (Herod et al., 2021: 428). In the UK, the industry contributes to around 3% of GDP and 5% of employment only but is less subject to seasonality. The regional distribution of the industry is more uniform in the UK than in Greece, where tourism is concentrated in coastal areas and islands.
Other differences are that Greece exhibits lower turnover rates and is less reliant on migrant workers. According to OECD data, the share of migration in the hospitality industry is 19% in Greece and 25% in the UK (30% according to the UK labour force survey). Black and Minority Ethnic workers are both overrepresented in UK hospitality (17% compared with 12% in the rest of the economy) and receive lower wages (Resolution Foundation, 2020). In cities such as London, already by the early 2000s, the migrant labour force in the hospitality industry exceeded 60% (Lopez et al., 2019). In Greece, Kapsalis (2020) mentions an increase in the participation of migrants in hospitality and tourism from 11% in 2008 to 13% in 2018, but this likely underestimates the real figure also including the undocumented migrants. Regular migration in Greece is less culturally and ethnically diverse than in the UK, coming mostly from neighbouring Balkan countries, and probably less skilled than it happens in Britain with a high share of students and graduates in the sector, a factor that might explain the higher turnover rates in the UK (Janta, 2011). Related to turnover, flexibility, and incidence of students and other part-time workers, the age of the workforce is higher in Greece (Karamessini et al., 2007;Kusluvan, 2003).
As noted above, some national-level institutional convergence between Greece and the UK has taken place with the former shifting towards a more liberal model of employment relations since 2009. Industry-specific dynamics such as low pay, long working hours, low-skilled jobs, and lack of career progression are common but there are also differences in relation to migration, turnover, role of the industry in the international division of labour and the national economy, composition of the workforce and flexible working practices. Unions are more active and militant in Greece, willing and still capable of influencing sectoral developments in ways that contradict recent studies. We argue that this factor may reveal more in relation to whether the differences between the two countries have any impact on the levels and nature of precariousness. Thus, even though comparative employment relations point towards a reduction in the differences between the two countries' industrial relations systems (Karamesini and Grimshaw, 2017), sectoral features and different trade union legacies might create opportunities for diversity in relation to actors' strategies and institutional preservation/change at the sectoral level. To sum up, the two key questions of this research: 1) Is recent convergence in the employment relations systems of the two countries reflected in differences in the conditions and experience of employment in hospitality?
2) What are the impact of different national trade union traditions and sectoral features on the ability of unions to resist precariousness and organize workers?

Methodology and empirical research
This paper is based on research conducted during 2018-2019, consisting of 70 semistructured interviews with hospitality workers in the UK and Greece on a variety of employment contracts. The sample included full-time and part-time, permanent, seasonal, and casual workers in hotels, restaurants, and café-bars in several cities and resorts, balanced in terms of gender and reflecting the primarily young ages which compose the hospitality workforce. It consisted of a broad range of hospitality occupations, both front of house and back house, including cooks and kitchen porters, receptionists, housekeepers and cleaners, barmen and waitresses, tourist guides, floor supervisors, and lower-level managers. We adopted the 'typical case sampling' approach (Patton, 2002), where the individuals selected occupy positions that are commonplace but also used quotas for each occupational position and employment status to maintain balance. Interviews were conducted in Glasgow, Edinburgh, Manchester, and London and in Athens, Thessaloniki, Herakleion, and Rethymnon, but some interviewees also narrated their experience working in other cities and resorts. Out of the 70 workers interviewed in total, 10 were non-natives. Most participants were recruited in their workplaces or through union contacts or social networks and a snowball strategy was used after initial contacts were made. In the UK, recruitment of participants proved more difficult and in the second fieldwork round advertisements and posts in relevant social media groups were used and £20 gift vouchers were offered to interviewees as a compliment and remuneration for their time. Gift vouchers are more commonplace in the UK than in Greece and their use in research is an established practice. A methodological implication of this practice was the decreasing significance of the role of social contacts as mediators in the recruitment process. The Research Ethics framework was revised to incorporate the remuneration of a section of the UK interviewees, which however has not caused a substantial difference in the composition of the sample. In addition, 6 interviews were conducted in each country with trade union officials, labour activists, employers, employer representatives, and other researchers in their capacity as experts.
The content of the interview varied according to the profile of the interviewee but generally included both current and past working experience, worker understandings of labour rights, collective action, and the prospects of improving the terms of employment. With trade unionists and experts, interviews revolved around issues concerning the institutional context and its impact on shaping realities on the ground, and content analysis was used to supplement the insights gathered from the literature and secondary sources more broadly.
The interview transcripts were subjected to thematic analysis. The interview content was subdivided into broad categories such as 'working time and pay', 'trade unions and collective action', 'collective agreements and contracts', 'working conditions and experience' in search of common elements, analogous ideas, and general patterns. The data from the narratives was thus organized according to several distinct themes, subjected to further interrogation, and reviewed in the light of insights from secondary sources.

Converging trends in precariousness
Hospitality workers in both contexts faced similar work organization issues, in line with previous studies, including excessive workloads, intense work rhythms, low job control, and routine and monotonous tasks. Working hours were irregular and unsocialespecially among flexible workerswhile heavy workloads and stress were widespread in both countries. Feelings of low employment status and strong perceptions about lack of career development and recognition by management were attributed to the absence of empowerment and participating mechanisms. Most respondents from both countries stressed that work intensification had increased across the sector in recent years with employers demanding more work with less and workers increasingly losing pay and autonomy: It is a quite challenging and stressful environment. You must do too many tasks at the same time. So many responsibilities and not enough people to cover them. But we are not recognized enough. For instance, salaries are quite low and the benefits you get for extra work are not really compensating for your effort (Female, 25, non-British, receptionist).
We have lost many colleagues since the last economic crisis. We are half the staff we were once and because of the rise in customers and work demands, we are doing much more work with much less pay. Many of us don't take the four days off (agreement with the employer) per month and in August we take none (Female, 45, Greek, cleaner).
In both countries, there was also widespread use of flexible and atypical work in the form of part-time, and temporary contracts, including zero-hours contracts in the UK, through which employers sought to swiftly adjust labour power to the inherently fluctuating demand in the industry. It was evident that the flexibilization of the Greek labour market such as the relaxation of temporary agency work in recent years influenced the experience of hospitality workers and brought the country closer to the UK model: When I first started working in this place, I realized that my pay was much lower than some of my colleagues. I was on the lowest grade (NMW) and worked longer with no overtime pay. It was because I was a temporary agency worker, and they were permanent (Male, 31, Greek, pastry worker).
Since the abolition of the catering wages boards in 1993formal institutional mechanisms regulating pay rates in the absence of collective bargainingthe main regulatory mechanism adjusting low-paid workers' wage rates in the UK has been the National Minimum Wage instituted in 1998. For many hospitality workers (those over the age of 25 at the time of our fieldwork), the minimum wage in its current 'National Living Wage' form operates as the going rate rather than the wage floor. As shown by recent figures, the average wage for hospitality and tourism is still 25% below UK's national average and 5% below the gross weekly wage for retail (Economic Insight, 2019). As many as 70,000 workers in hospitality were paid below the minimum wage in 2017, while almost four-in-ten are owed wages for hours that they have worked (TUC, 2019).
In Greece before the 2008 crisis and due to a more regulated labour market framework and sectoral agreements, wage levels were significantly above (on average 20% above the NMW) despite variation in terms of seasonality, size, and scope of services offered (OMED, 2020). However, following the reduced national minimum wages introduced in 2012 as part of 'austerity', the sectoral agreement predicted wage reductions of 15% (OMED, 2020). For this reason, significant similarities are observed between the two countries in recent years in terms of remuneration as the quotes below illustrate: Unsocial and overtime were fully paid, so my salary could reach €1,300-€1400 per month if I worked those hours. The same applies to the number of rooms I had to clean…Things have changed dramatically, since the abolition of the sectoral agreement my salary is not far from the national minimum wage (Female, 47, non-Greek, cleaner).
Most employees get the national minimum which is £7.83. I don't know many people that get more than that. We don't get extra pay for overtime work so when calculating the hours worked sometimes, we get paid less than the national minimum wage. But this has been the culture in this industry for many years (Female, 45, non-British, cleaner).

Diverging trends in precariousness
In the UK food and accommodation sector, 23.7% of the workforce is on zero-hours contracts, while part-time work comprises 37% of total employment, substantially higher than the 24% economy average (ONS, 2019). In Greece, part-time employment stands at 20% (national average at 27%) of all employment in the sector, although it has risen significantly since 2010 (Insete, 2020). Institutional factors such as stricter Greek labour regulation of flexible work in hospitality (and the economy as a whole) before the crisis account for this difference, along with the different structure of the Greek industry that relies more on seasonal work and thus temporary but full-time contracts. Despite the partial convergence mainly due to the recent relaxation of Greek employment law promoting agency work and outsourcing as a mechanism for reducing costs as well as facilitating greater flexibility compared with permanent directly employed staff the use of flexible work is more widespread in the UK.
According to UK trade unionists, this is facilitated by the more liberal institutional context, the larger size of hospitality businesses, and the absence of workplace unions to oppose such practices. In Greece, the picture is more complex and varied, with some path dependency still evident despite the recent convergence of the national systems of industrial relations. For instance, according to the sectoral agreement, which was maintained after 2012 despite significant pay reductions, seasonal workers have the right to be recruited (priority in the case of re-calling of seasonal workers year after year) by their employer again, a practice that does not exist in the UK (Papadopoulos and Lyddon, 2020). Despite the dismantling of collective bargaining in Greece the two countries continue to diverge in terms of wage inequality: in Greece, the percentage of low-wage workers was 37.21% in 2018 (from 46.69% in 2014) while in the UK was 65.89% (from 54.56% in 2014) (Eurostat, 2018).
The different place and role hospitality has in the two economies can also explain the different attention it gets from trade unions which in Greece are ready to put more resources and effort into union organizing activities. According to a trade unionist interviewed, the increased recruitment of primarily young workers and their accumulated experience of poor working conditions in hospitality have helped the development of collective organization and activity, with the sector seen as less of a transient location, and consequently, grievances have substantially increased in recent years. The strong growth of hospitality and tourism in many regions also means that contrary to the UK, where most actions are concentrated in London, unions are present and active in large hotels in popular tourist destinations. Many demonstrations, strikes, and rallies were coordinated by hospitality unions across Greece, creating pressures on employers and the State (Ertnews, 2020). During 2017-2020 Greek sectoral unions went on strike five times with three of them being held during the high season (July 2017 and 2018 and June 2019) while during the same period many hospitality unions mainly affiliated with the class-oriented Hotel Catering Union of Attika (HCUA) held numerous militant protests and demonstrations in relation to work intensification, flexible contracts and collective bargaining rights (Dermani et al., 2021).
Although UK hospitality employs a significant number of young people, migrants, and women, it never acquired a central role in the national economy and many workers saw their jobs as short-term and entry-point ones. That might explain why in a context of high labour turnover, extreme flexibility, and lack of experience with unions among young workers, Unite, the only UK union truly active in hospitality, has invested little in organizing the workforce (Wood, 2020). The quote below is illustrative of the effects of these factors on unions' limited presence on workers attitudes: I don't think it has any effect or I feel it might have negative effects if they found out I was in a union. But you know, the unions are not that visible. At least I didn't see them in my experience in this hotel (Male, 29, non-British, receptionist).
The higher proportion of migrant workers in the UK explains why unions have faced greater difficulties when attempting to organize hospitality workers, especially in places where migrants dominate. Contrary to Greece where migrant workers are less visible because they engage in seasonal, informal work, migrants in the UK have more formal employment relations despite being at the lowest end of the hospitality labour market facing more insecure employment, low pay, and longer working hours than native workers. UK unions have experimented with community organizing which has generated some public awareness of hospitality employment conditions but in the main have failed to produce any sustained organizing imprint in a context of strong employer anti-union practices (Wood, 2020). This exacerbates fear and isolation which together with social norms like accepting bad working conditions and not complaining makes migrant workers rather vulnerable and reluctant to resist: We are scared to put a formal complaint, we don't know what will happen after, we have families, I cannot afford to be without work since I am a foreigner in this country, and I don't have the network to support me in case of dismissal. So, no I wouldn't join a union even if there was one (Female, 39, non-British, room attendant). Also, the low attachment to the labour market and the transient-liminal employment statuses among migrants in the UK (especially young and educated ones) create high fragmentation, and uncertainty, thwarting workplace solidarity and sustained union presence (Jordhus-Lier, 2015). The generally large business size in the UK facilitates less direct and more short-term and impersonal forms of employmentwhen involving outsourcing and agency work, it fragments further the workforce raising additional obstacles to collective action: Hotels and restaurants are property-based franchises belonging to hedge funds and private equity firms: they own the brand and the building but have little concern for employees; the shareholder comes first. Most don't recognize or actively oppose unions and workers are rather vulnerable and fearful to resist work degradation and low pay (Trade unionist, UK). Different ideological orientations and logics of action adopted by Greek and the UK unions explain some of the observed differences. Due to historical reasons, sectoral unions in Greece, such as the Federation of Panhellenic Federation of Food Workers -Tourism are ideologically and politically divided with some of the fractions, maintaining class traditions within the labour movement (Karakioulafi et al., 2021). So, although factionist fragmentation can be a weakness, yet at the same time, it also fosters internal union debates and supports the expression of different union ideologies. The presence of more militant and class-based unions often politicizes disputes with employers over workplace issues, leading to their framing under 'a contestation conflict of interestsoriented' discourse that promotes solidarity among various other unions from other sectors. In contrast, reflecting the tradition of market union identities, Unite is more concerned about economic and employment issues, framing resistance within a minimum protection of employment rights discourse as evidenced by the request to introduce the 'true' Living Wage and the recognition that less precarious work can help hotels increase job retention and productivity: The living wage would solve some problems of the industry like high staff turnover and skill shortages that the sector has suffered for years. It is also the right thing to do (Trade unionist, UK).
Combined with agentic action underpinned by militancy and activism, the sectoral legacy of the past in Greece, based on the tradition of collective agreements, permanent contracts, and some employment rights, was utilized by unions to reverse some of the Troika imposed reforms such as the subminimum wage and force employers to accept some demands with mechanisms such as signing agreements with 'association of persons instead of trade unions falling into disuse. By 2018, the erga omnes effect of the sectoral agreements returned, arresting their erosion. Union strength at the workplace level was important in challenging employers' unwillingness to implement collective agreements and significantly changed the course of events confirming the power of agency in a context where some path dependency enables strategic action by actors: We gave a very tough battle for sectoral agreements in the hospitality industry. The bigger struggle is to force employers to implement them not only at the sectoral level but in every workplace and this has been only possible only where workplace unions are strong. Otherwise, employers take advantage of various legal clauses and the non-existence of Inspections to avoid the agreements (Trade unionist, Greece).
Workplace union presence has been built upon a legacy of successful confrontation with employers and the State and is praised for its active involvement in preventing large hospitality businesses from liberalizing the labour market on the pretext of the 2004 Olympic Games (Ziogas, 2012). This legacy continued and manifested itself in various cases amid very unfavourable conditions. HCUA, a class-oriented union prevailing in Athens, made several unannounced workplace interventions forcing on some occasions firms to respect employment rights and pay wages on time. Greek unions have been able in several cases to prevent employers from taking advantage of legal-institutional changes. The most recent example is the successful opposition to the introduction of the gig economy model in food delivery with mobilizations enjoying significant public support through social media and resulting in both securing regular contracts and the setting up of a workplace union (News 24/7, 2021). The role of HCUA, its ideological stance, and the legacy of permanent contracts and limited flexible work were central in this: An important element of our success was that we chose not only to oppose the introduction of the gig economy model but also to fight for the conversion of temporary contracts into permanent ones. This is a set of demands that we had developed long before the specific mobilization and was accepted by workers (many of which had permanent contracts in the past) because it responds to their needs which are in opposition to employers' profit-seeking and employers-friendly unions that don't oppose flexible work (Leader of the HCUA union).
Militant action kept the political pressure high and contributed to the reaffirming of the mandatory character of the sectoral collective agreement by the Supreme Labour Council in May 2021. In another case, a local militant union successfully stopped the introduction of agency work in a large Athens hotel by mobilizing work stoppages of cleaners in 2018, forcing the management to hire staff on permanent in-house contracts instead. Such actions boost the image of unions among non-unionized workers: I think that unions are positive to be honest. I know that prior to the crisis my parents' generation had some employment rights because of unions and it feels like my generation needs to also do something to fight precarious work and low pay. I recently joined a union myself to fight against the working conditions in my workplace (Male,21,Greek,houseman).
On the other hand, the lack of experience with collective organizations and socialization in an individualized and flexible world of work limited the legacies UK unions could draw on. As a result, most attempts to organize hospitality workers achieved little, with unions unable to secure recognition and increased density. Especially in non-union workplaces, workers displayed fear of victimization for pursuing collective action since employers view the demanding of rights as 'bad' behaviour that could lead to dismissal (Guardian, 2015). This led many to express ambivalent stances towards unions, summarized in the view that collective action can be positive in general but might not be such a good idea in their own workplace. Seasonality and temporariness of employment were also important factors in both cases: For me in the UK, I don't think it will have any effect if I join any union, I don't see the advantage of joining unions in the UK unless I was in a different sector (Male, 29, non-British, receptionist).
During the crisis years, the increasing recruitment of agency and outsourced workers who have no affiliation with unions, no memory of industrial action, and short-term vision worried trade unionists and labour activists. Seasonal workers who tend to be young and migrants have also significantly increased and this demographic factor, along with the lure of tips which in some settings can be sizable, and their eagerness to earn as much as they can, make attraction to unions less likely (Lucas, 2009).

Discussion and conclusions
Our research focused on the extent and the ways in which employment conditions in hospitality are converging towards a low-road and low-pay model, and on the role of trade unions in it. Our findings suggest that hospitality workers in Greece experience similar conditions to their UK counterparts due to institutional erosion, which enabled employers to adopt more flexible practices to respond to fluctuating demand. Common precarization patterns were observed in terms of the quality or experience of work, whereby fixed term/ seasonal, zero-hours/on-call, casual, agency, and even self-employment contracts are common in both countries since 2010 as Greece moved closer to a liberal, voluntarist model of employment relations. Similarly, poor working conditions such as work intensification, low pay, and violation of employment rights were found in both cases, despite differences in extent and degree.
While employment protection for most hospitality workers at least is insufficient, as rights enforcement encounters numerous difficulties, there is a dense fabric of laws, rules, and regulations in place in Greece which partially survived the 2008 crisis and the legal reforms that followed, and which can be mobilized. Sectoral regulation is important here as some rights including the right of temporary workers to be re-hired are established in the sectoral collective agreement, while flexibilization remains even after the liberalizing reforms during the crisis, less advanced in Greece. Resistance from unions and sectoral collective agreements in hotels have constrained to some extent the use of flexible contracts of employment.
The regulatory tradition in Greece would not be sufficient to prevent precariousness if sectoral characteristics and legacies did not allow unions to oppose employers' practices and resist deregulation. Specific sectoral features of the Greek case including the significance of the industry, limited formalized migrant workforce, and longstanding union presence allowed unions to thwart national measures and precarity-enhancing trends to a much higher degree compared to the UK. The intersection between path dependency and agency was indicative in the demands that unions have put forward which include hardwon rights of the past that were jeopardized since austerity and labour market reforms were enacted. These include permanent contracts, pay rises, collective agreements, and special insurance against the unhealthy and dangerous conditions of the industry.
Unions managed to utilize these features because of the legacy of militancy and classoriented union identities within the Greek labour movement that on many occasions in the past opposed employers' and the State's aggressive practices and won concessions. Recent cases of successful mobilizations and strikes confirmed the legacy of this type of unionism and its ability to use power resources as a means for institution-preserving as in the case of implementing sectoral agreements where unions were strong (Frege and Kelly, 2020). So, even though the institutional change was very pervasive in Greece, building on their specific ideological and power resources, actors managed at least to some extent to prevent the internalization of the new employment order and adopt agentic actions that rebel against rule-following (Streeck, 2011: 12). We argued that opportunities for actors to break with some elements of the new liberal employment landscape were linked to inherited normative and ideational traditions (Frege and Kelly, 2020: 17) within the Greek labour movements that enabled unions to defend their role. The maintenance of sectoral agreement and associated rules, despite their weakening, and their being alive in the memory and experience of workers allowed trade unions to use these as resources in their resistance even in very adverse institutional conditions.
In the UK on the other hand, past legacies intersect with sectoral features producing negative outcomes for workers including high precariousness and limited resistance. The less significant positioning of the sector and the recruitment of more migrants in comparison to Greece, combined with the utilization of flexible practices created a very temporal employment landscape with high turnover that discouraged the dedication of union resources as indicated by the low organizing attempts made by hospitality unions. Union absence in UK hospitality due to a very hostile legal framework, employers' anti-union practices and decentralized models of management increase the levels of isolation and fear and reduce the confidence of workers in collective action. However, as recognized by workers and trade unionists alike, the dangers of liberalization/ decentralization are not negligible in the Greek case either, since the younger generation of workers has limited contact with unions and is unaware of employment rights of the past that in most non-union workplace are absent.
The article highlighted that although opportunities for resistance to precariousness are uneven and unstable, under certain conditions unions can mobilize existing historical and institutional resources to achieve concrete improvements for workers. Our findings challenge the claim by previous studies such as Baccaro and Howell (2017) that neoliberal convergence towards poor and flexible working conditions is universal in low-paid sectors. Examining the same sector in two national contexts, we identified that the legacy of strong and active unions is a significant factor that can both complement and supplement path-dependency reducing precariousness and improving to some extent the experiences of workers even at the low-paid end of the economy. Our contribution moves comparative studies in low-paying sectors a step forward by emphasizing the need to take more seriously the significance of power resources, sectoral context, and bottom-up union actions even in situations where national institutions have been seriously undermined and capital-friendly policies have been implemented.