Corporate Social Responsibility in Temporary Agency Work: A Study of Restaurant Work in Finland

Ever expanding use of temporary agency work (TAW) easily attracts irresponsible actors on the business field in the hopes of making big profits and many narratives frame the emergence of contingent agency work as a positive evolution for employers and workers alike. The purpose of this study is to examine corporate social responsibility (CSR) in temporary agency work (TAW) in a small restaurant context to consider what kinds of challenges and opportunities can be identified from the perspectives of the workers, agencies, and user companies. A directed content analysis method was used on semi-structured interviews collected from restaurant workers, managers, and temporary agency experts, including a pension insurer and a labour union in Finland. To gain a new holistic understanding of this phenomenon, the pressures–disorganization–regulatory failure (PDR) model was used as a research framework. The CSR challenges identified were multifaceted, and conflicting views were common. A wide variety of CSR challenges and evolving strategies were found to avoid TAW regulations.


Introduction
Temporary agency work (TAW) has established itself in various business branches, and many actors have entered the industry, not all of which strive to act responsibly (Forde & Slater, 2005;Knox, 2018). An expanding market easily attracts irresponsible activities in the hope of making big profits (Knox, 2018), and many narratives frame the emergence of contingent agency work as a positive evolution for employers and workers alike (Robinson et al., 2019). For instance, it has been debated whether a temporary contract will make it easier to obtain normal permanent employment (Chambel & Sobral, 2019;Wears & Fisher, 2012) On the other hand, agency workers often work for the same company for extended periods of time, without getting normal permanent employment, meaning that the agencies constantly exceed the maximum time set for temporary agency work for one employer (Bednarska-Wnuk, 2015). TAW has been criticized for delegating decision-making power between an employee and an employer to an intermediary, the temporary work agency (Felfe et al., 2008;Wears & Fisher, 2012). Many researchers from various countries have reported greater unfair treatment and situations for people working in TAW compared to permanent and full-time workers (Akkerman et al., 2020;Cajander & Reiman, 2021;Grimshaw et al., 2019;Håkansson & Isidorsson, 2012;Kauhanen & Nätti, 2015;Maroukis, 2016). The arguments for and against TAW are of many kinds, reflecting the various viewpoints. For example, the commonly used argument on the flexibility of TAW looks quite different from the workers' and user companies' perspectives (Bednarska-Wnuk, 2015;Forde & Slater, 2005;Hveem, 2013;Koutsimpogiorgos et al., 2020;Kost et al., 2020;Robinson et al., 2019).
Earlier research on corporate social responsibility (CSR) in TAW context has shown the fragmentation of CSR concerning agency workers, agency workers suffering more drawbacks than normal employees, having less influence on work-related decisions, and receiving less support from co-workers (Bednarska-Wnuk, 2015;Akkerman et al., 2020;Macke & Genari, 2019). Also, processes, practices, and practical CSR arrangements for occupational health and safety in TAW have been under critical debate, and studies such as those of Lippel et al., (2011) and Vanroelen (2019) have highlighted the challenges of TAW in this context. At worst, it has been reported that the user company can circumvent protective legislation designed to ensure occupational health and safety and a proper income for agency workers (Lippel el al., 2011;Strauss-Raats 2019). Even though this has been known for years (since the beginning of the 2000s), it seems that very little improvement has been achieved in this area (Forde & Slater, 2005Knox, 2018). Moreover, research conducted in this area has identified new and emerging ways to avoid regulations and evade CSR in TAW context (Bednarska-Wnuk, 2015;Forde & Slater, 2016;Fostervold et al., 2018;Knox, 2018) even when responsible business practices and the management of sustainable development have been identified as important elements of successful company performance (López-Pérez et al., 2017;Malik, 2015;Rodriguez-Gomez et al., 2020).
Further research has been called for, both on the manifestations of good CSR and the incentives to circumvent regulations (Forde & Slater, 2016), including across different national contexts (Knox, 2018) and focusing on small and medium-sized companies (Ortiz-Avram et al., 2018;Stekelorum, 2020). So far, the work-related factors and employee reactions to CSR have received less research attention (Gaudencio et al., 2020;Hakanen et al., 2019;Robinson et al., 2019). Furthermore, future research is needed on productivity in TAW in relation to CSR, examining, for example, job requirements, insecurities, and commitment to work (Burgess & Connell, 2006;Cajander & Reiman, 2021). This study aims to address CSR in a typical TAW context, namely restaurant environments. Restaurant work is labourintensive and associated with a variety of different risks from the CSR perspective. Despite a variety of large commercial actors, like chains operating in this field, most of the restaurants are small and privately owned businesses. For instance, in the European Union area nine out of ten companies are classified as small or medium sized companies (Fooddrink Europe, 2017). To gain a new holistic understanding of this phenomenon, the pressuresdisorganization-regulatory failure (PDR) model (Bohle et al., 2015(Bohle et al., , 2017Strauss-Raats, 2019) is used as a research framework. Exploring TAW within this framework enables a multi-faceted analysis covering both the social and economic sides of CSR. A PDR analysis of TAW allows us to answer the following research questions: RQ1: From the PDR-model perspective, how is CSR realized in temporary agency work in the restaurant context? RQ2: Considering this PDR model-oriented examination, what kinds of challenges and opportunities can be identified from the perspectives of the workers, agencies, and user companies involved in TAW in restaurants?

Background
A well-established PDR model (Bohle et al., 2015(Bohle et al., , 2017Strauss-Raats, 2019) serves as the analysis framework for this study on CSR in TAW. TAW is one emerging form of nonstandard employment in the ever-changing world of work (Iavicoli et al., 2018;ILO, 2021). Complex ethical dilemmas concerning new forms of work in general (e.g., Iavicoli et al., 2018) and in TAW (e.g., Knox 2018;Oyetunde et al., 2022;Robinson et al., 2019) have been highlighted in the literature. In this study the focus is on the complex demands CSR has brought along to TAW. The following sub chapters provide necessary background for CSR in TAW and for the PDR model.

CSR in TAW
TAW is characterized by a triangular working contract. This means that the work employment relationship includes three parties: the user company that uses the workforce, the temporary work agency that supplies the workforce and is the worker's official employer, and the worker who performs the work tasks (Hünefeld et al., 2020;Knox, 2018). TAW is used for many reasons, from the need to fill positions during labour shortages to short-term pressure to lower labour costs (Burgess & Connell, 2006;Forde & Slater, 2005), or to restructure and relocate human resources (HR) management elsewhere (Townsley & Stohl, 2003), and in extreme cases abuse of loopholes in legislation (Knox, 2018). To ease the availability of the workforce, user companies may enter long-term strategic partnerships with temporary work agencies (Burgess & Connell, 2006;Forde & Slater, 2016). These strategic partnerships allow user companies to reduce risks associated with market uncertainty and labour market challenges as well as avoid CSR (Burgess & Connell, 2006;Forde & Slater, 2016;Knox, 2018). TAW is a managerial and CSR challenge as agency workers must conform to the expectations and standards of two employers, consequently blurring employment conditions and the implementation of the employment contract (Johnstone & Quinlan, 2006;Knox, 2018). Furthermore, managers of user companies often regard the management of agency workers as not being their matter but that of the agencies (Burgess & Connell, 2006). For this reason, user companies may differentiate their HR practices for the different worker groups, with their own workers receiving greater investment and corporate commitment (Burgess & Connell, 2006;Schmidt et al., 2018). Temporary agency workers have been identified as having less control over their working hours and other aspects of work than permanent workers due to their weak labour market position and contingent employment status (Hünefeld et al., 2020;McNamara et al., 2011;Oyetunde et al., 2022). In turn, managers of user companies need to manage the organizational and stakeholder trust damage, for example high turnover, that can happen due to the presence of insecure and uncommitted agency workers (Bednarska-Wnuk, 2015;Burgess & Connell, 2006;Cajander & Reiman, 2019;Pirson & Malhotra, 2008).
CSR is a globally understood strategic initiative that seeks to cover the social, environmental, and economic dimensions in the actions of organizations, with the aim of achieving ethical and sustainable development together with the stakeholders (Abriata & Delautre, 2020;Bednarska-Wnuk, 2015;López-Pérez et al., 2017). Especially in TAW some companies view socially responsible activities as an unnecessary burden because of the costs and the complex nature of the potential benefits and instead are focusing on economic and environmental sustainability, corporate reputation, and brand image, while minimizing the social dimension, including well-being at work (López-Pérez et al., 2017;Robinson et al., 2019). This has brought forth the need to pay attention to social and economic matters and the development of human capital in TAW from the viewpoints of both the company and the worker (Macke & Genari, 2019). However, these viewpoints are often contradictory in practice (Macke & Genari, 2019). Authors like Riera & Iborra (2017) have discussed this ethical dilemma from the perspective of the corporate social irresponsibility (CSIR). Accordingly, CSIR is opposite to CSR as it focuses on the actions that can be considered irresponsible from the viewpoint of the environmental, economic and social sustainability. (Riera & Iborra 2017, Carroll & Brown, 2018).
CSR at the individual level should be regarded involving working conditions that promote long-term physical, social, developmental, and economic effects for employees (Dellve & Eriksson, 2017;Fostervold et al., 2018;Macke & Genari, 2019). At the organizational level, responsibility should not only include social, economic, and environmental concerns inside the organization but also extend these concerns to cover all interactions with all stakeholders (Dellve & Eriksson, 2017;Fostervold et al., 2018;Hakanen et al., 2019). One threat to CSR, especially in a TAW context, is the principle of using and discarding workers if there is a constant supply of new workers (Fostervold et al., 2018). If sustainability is not taken care of all work will be less stable and more difficult to obtain and maintain in the future (Fostervold et al., 2018;Knox, 2018).
During the past few years, European countries and organizations have sought to implement regulatory requirements to restrict or impose conditions on the use of agency work (Alsos & Evans, 2018;Buchanan et al., 2013;Johnstone & Quinlan, 2006). These measures have had only limited effects, and agency workers are still vulnerable to exploitation, such as lower salaries, fewer overtime entitlements, and employment relationships that are outside the existing labour law (Alsos & Evans, 2018;Cotton, 2015;Forde & Slater, 2016;Johnstone & Quinlan, 2006;Knox, 2018). TAW is evolving in many ways and bringing challenges to CSR. For instance, Forde & Slater (2016) have highlighted the challenging nature of the increasingly complicated and constantly changing contract arrangements between agencies and user companies, the changing regulatory environment, and the changing role of agencies in salary setting. As the most recent development, user companies have been criticized for their attempts to avoid paying employee benefits (e.g., superannuation, sick leave, and annual leave) by hiring workers under commercial contracts rather than as employees (Forde & Slater, 2016;Knox, 2018). Furthermore, investments in personnel training are frequently avoided by using sham contracts (Forde & Slater, 2016;Knox, 2018).

Pressures-disorganization-regulatory failure (PDR) model
The pressures, disorganization, and regulatory failure (PDR) model offers a holistic framework for viewing the realization of CSR in the contexts of TAW (Strauss-Raats, 2019) and the restaurants (Bohle et al., 2017). The elements of PDR help explain inferior outcomes in contingent work and develop policies and practical interventions to improve work-related outcomes (Bohle et al., 2015;Strauss-Raats, 2019).
In the PDR model, economic pressure relates to insufficient or irregular income due to issues such as long or irregular hours, contingent wages, multiple jobs, and unpaid sick leave (Underhill & Quinlan, 2011), which have been identified as impairing workers' financial situation and lowering their work motivation (Bohle et al., 2015(Bohle et al., , 2017Strauss-Raats, 2019;Underhill & Quinlan, 2011). In the PDR context, the focus is on the pressure caused by financial burdens and payment systems rather than the quality of various types of rewards, such as promotion and plaudits (Bohle et al., 2017).
Disorganization in the PDR model refers to fractured work practices due to low employer commitment to the contingent workforce (Strauss-Raats, 2019; Underhill & Quinlan, 2011). It is characterized by interference and inefficiency arising from inefficient communication and procedural errors between different groups in the workplace (e.g., between permanent employees and agency workers or between workers and management; Bohle et al., 2015;Bohle et al., 2017). Other elements of disorganization include the excessive use of inexperienced or poorly trained workers and ineffective induction, training, and supervision, which lead to unfavorable work outcomes (Bohle et al., 2015(Bohle et al., , 2017. Disorganization also relates to limitations on the ability of workers to organize to protect themselves, especially those organizationally or physically isolated from co-workers (Bohle et al., 2015). TAW is particularly susceptible to disorganization, as its triangular nature increases the probability of employer responsibilities falling between different employers (Strauss-Raats, 2019).
Regulatory failure concerns inadequacies in the application of minimum labour standards. This includes weakened employer compliance in the allocation of employer responsibilities and obligations, the failure to monitor and enforce labour law, and weaknesses in the protection of non-standard workers (Bohle et al., 2015(Bohle et al., , 2017Strauss-Raats, 2019;Underhill & Quinlan, 2011). It also involves restraints on workers' knowledge of relevant rules and standards and their willingness or capacity to report problems (Bohle et al., 2015). In previous research, there have been clues regarding interconnections and overlaps between the elements of the PDR model, with workers' financial and economic pressures being dependent on employer practices and both being based on regulatory failure (Strauss-Raats, 2019).

Data collection
This study was a part of a large national research project focusing on sustainable gig work in Finland. Researchers from four universities participated the project. The project plan including descriptions for data collection and analysis was approved in Fall 2020 by the Human Sciences Ethics Committee of the Helsinki Region Universities of Applied Sciences. In the project plan, the principles of research ethics were followed based on the Finnish National Board on Research Integrity and the General Data Protection Regulation (TENK; EU, 2016/679).
Empirical data were collected through semi-structured theme interviews during January-February 2021, with questions based on the widely used Conditions for Work Effectiveness Questionnaire-II (CWEQ-II; Almost & Spence Laschinger 2002;Spence Laschinger et al., 2001). In addition, questions about the state of TAW and its development were asked. The CWEQ-II measures access to different structures of opportunity, including the possibilities for organizational mobility and learning, obtaining knowhow for efficient work, receiving feedback and guidance from superiors, peers, subordinates, and outside parties, and having the economic means, materials, time, and equipment required to do the work. The CWEQ-II also measures the structure of two kinds of power: formal power and informal power. Formal power is derived from job characteristics such as flexibility, adaptability, creativity, decision making, visibility, and importance to the purpose and goals of the organization. Informal power is derived from social connections and the development of communication and information channels (Almost & Spence Laschinger, 2002;Spence Laschinger et al., 2001). The themes of CWEQ-II are compatible with the PDR model and similar methods have been applied before with the PDR model (Strauss-Raats, 2019), for that reason, it was chosen for application and use in this study. However, the research group decided to apply the questionnaire in an interview format because they considered that a survey was not likely to produce answers in the detail needed for this study due to the challenging nature of TAW.

Participants
Prior the interviews, the project group discussed the inclusion criteria for the interviewees. The project group agreed to select informants that have practical experience on TAW either from the agency worker, the user company or the agency perspective. In addition, other stakeholders were considered when they had extensive professional experience on the topic. The first author of this study contacted potential user companies and agency workers using his/her professional networks (including employee unions). To find potential interviewees from the agency and other stakeholder perspective, the author group consulted the steering group of the project. Representatives of the employee and employer unions, in addition to authorities participated the steering group. Potential interviewees were contacted by phone. During the first contact the interviewees were provided a brief introduction of the study and briefly introduced the inclusion criteria.
All interviewees were provided a written consent prior the interviews. In the consent the interviewees were notified formally of the study purpose. As discussed in the consent, participation was voluntary. Hence the interviewees were informed that they can withdraw whenever they want without any consequences. In all, 29 people in Finland were interviewed. The interviewees represented restaurant managers (n = 5), restaurant agency workers (n = 7), and experts in TAW (total n = 17). The TAW experts represented a national labour union for people working in the private service sector (n = 1), a national pension insurer (n = 2), and temporary work agencies (n = 14). The restaurants where the interviewed managers worked were privately owned, had operated for over 10 years, and provided both food and alcohol service. The restaurants were small by employee count (between eight and 20) and had a light organizational structure, with a manager, a pair of shift managers, and five to 17 permanent full-time workers. In addition, the quantity of agency workers varied by week, ranging from zero to six. These restaurants were selected for study because of their long period of operation and their regular utilization of TAW. All agency worker and manager interviews and one expert (employee union) interview (n = 13) were conducted as personal interviews where one to two researchers participated. Seven agency and one expert interview were conducted as group interviews each having two interviewees. On average, the interviews lasted for 60 min. All interviews were recorded and transcribed. In addition, written notes were taken during the interviews to support the analysis.

Data analysis
Interview material was analyzed using NVivo software (Jackson & Bazeley, 2019) by means of directed content analysis (Hsieh & Shannon, 2005;Assarroudi et al., 2018). The PDR model (Bohle et al., 2015(Bohle et al., , 2017Strauss-Raats, 2019) directed the coding of the data. During the analysis, the main author reviewed, coded, and immediately analysed the data. Following the Directed content analysis guidelines by Assarroudi et al., (2018), each interview transcript was carefully read through several times to acquire deep understanding of the data. In the following phase, significant statements were highlighted with colour coding using the NVivo software. These statements were then categorised according to the PDR model to answer RQ1. At the final stage of the analysis, the author group jointly discussed the findings and their relations between each other to answer RQ2.

Results
The following results are structured by the perspectives of the PDR model, following the categorizations of Bohle et al., (2017) and Strauss-Raats (2019). Accordingly, economic pressures, disorganization, and regulatory failure relating to TAW in restaurant work are examined. The findings presented are intensified with interview quotations. The quotations were translated from Finnish to English by the first author.

Contingent working and payment
The interviews showed that agency employers in general favoured full-time and long-term employment relationships for their listed workers and tried to minimize workers' disadvan-tages in terms of irregular income and hours, contingent wages, and time needed for preparation of work in the case of multiple jobs. In this way, the agencies received maximum benefit for themselves, as they did not get paid if their workers were not employed working for the user companies. It was observed that the longer the contract was valid with the user company, the easier and more profitable it was for the agency, as caring for the worker becomes less labour intensive, but the money keeps coming in. Therefore, the evidence suggested that the agencies did not in general discard their workers. However, this emerged as an issue in some user companies that were trying to save costs with TAW or trying to get rid of a troublesome worker. In some cases, the user companies were highly satisfied with the agency workers' performance and were even willing to hire the worker as their own to avoid remuneration to the agency. Letting a profitable worker go was considered being against the interests of the agency, so they often had a contingency in place to prevent a worker leaving. The agencies had formed contractual arrangements where fines were set for the user company should they try to hire a worker as their own. From the CSR perspective, this is somewhat contradictory and debatable, as it takes away workers' right to decide for themselves, but on the other hand, from an economic perspective, the absence of any contingency agreement would eventually cause harm to the agencies' business. Agency Expert 1 expressed the pursuit of continuing relationships in the following statement: After all, we will not earn a single euro unless we get people to work. It is in the interest of all of us to serve these jobseekers best of our abilities, so they get a suitable job for themselves and work there for a long time. Then we will have better customer channels.
In the interviews, it was highlighted that there were numerous entrepreneurs in the temporary work agency business, as it was very easy to start a new agency, with only minor investments required. Therefore, agencies saw that the competition was hard, and there was huge variation in professional practices. Agencies emphasized that the intensifying competition established pressure for responsible conduct, as it was easy for user companies to simply change the workforce supplier if the agency did not comply with their demands. These challenges were well known to all parties of temporary agency work contracts, and, as a solution, parties often entered long-term partnerships. One aspect sought after in partnerships was predictability. When agencies and user companies knew each other's timetables and deadlines, they could more easily predict the need for labour and make sure the labour was available at the appropriate time. This conduct also benefited workers, bringing predictability and certainty to their lives. Another aspect mentioned by user companies was outsourcing, as they liked to focus on their core business and let someone else take care of HR functions. Many user companies did not have the time or expertise to handle all the agency workers periodically working for them.
One contract-related theme raised in the interviews were worker pools, which it is particularly important to consider in the light of CSR. Worker pools are groups of workers reserved by the agency to be used in the specific user company at the appropriate time.
Workers who have experience working for the user company or have been inducted are kept in a worker pool. When the need arises, the user company then takes workers from the worker pool, and after the job, the worker goes back to the pool. In this way, user companies always have workers ready to come to work. From the viewpoint of CSR, this situation can be challenging, as it is not always clear whether the worker can work anywhere else while waiting for job opportunities in the worker pool. Some workers assigned to the pool highlighted how they did not get work from the agency anywhere other than in the same user company, while other workers listed by the agency could work anywhere. Workers experienced this as increasing work-and income-related insecurity and reducing their commitment and effort. Fortunately, the workers reported that this situation had improved after their complaints to the agency. Agencies emphasized the importance of long-lasting partnerships, stating that there have been ethical considerations for keeping workers in stock, while user companies were satisfied with always getting workers. Agency Worker 1 provided their view on this situation: I have toured workplaces of this same company for a while now, and it seems that I do not get work shifts anywhere else even if I have asked, and others [listed by the same agency] have gotten them.

Pressure to work irregular hours
Flexibility was found to be a central element in the marketing of agency work as well as something that was sought after by the workers, but this flexibility did not come true without certain challenges. In an ideal situation, flexibility was regarded to follow a smooth practice whereby the employer offered work shifts and the worker was then allowed to freely decide whether to accept these offered shifts. However, in practice, this idealistic arrangement was being strongly influenced by economic pressure: the user companies were simply not being able to tolerate production interruptions, so the workers had to control their labour availability together with the agencies using flexible work arrangements. This economic-centred viewpoint, offering work when workers were needed quickly and then dismissing them when the need receded, was a source of dissatisfaction expressed by the workers in the interviews, as highlighted in an interview quotation by Agency Worker 2: The work situation is not so good that the worker can really choose. "I will do that, and I will pick that." That is a very rare situation. Moreover, it is like constantly looking at a mobile phone or a computer: are there work shifts available to take before someone else takes them?
The workers were in a decision-making position in which they were freely able to choose their workplaces and working hours in the case of only a few occupations. Highly esteemed cooks were mentioned as an example of such occupations.

Fractured occupational health and safety arrangements
Differences were identified in the occupational health (OH) arrangements in terms of the practical division of responsibilities between the agency and the user company. The agency, as the primary employer, had the legal responsibility to organize OH services for the worker. However, the agencies felt that they had to bear the economic risk by themselves, as such costs were not easily shared with the user companies. The user companies thought it was easy to change agency if additional costs piled up. In some cases, the user companies extended their OH coverage to the agency workers they used. However, even in these cases, the details of the OH arrangements were insufficiently communicated to the agency workers. The multifaceted challenges related to OH arrangements emerged clearly in an interview with Restaurant Manager 2: The agency from which we receive workers would like to get OH in the cheapest way. If one of our agency workers falls ill, we force the agency the worker comes from to pay for a visit to the doctor. Our ideology is that when occupational health care is functioning well, the happier the worker is and consequently the better is the customer service. Due to this, our agency workers also have very few sick leaves. The agency has noticed that it is sensible and in its best interests to pay if a worker needs a doctor's visit.
Matters of occupational safety were an obligation for the user companies. The agencies had very limited possibilities to affect the promotion of occupational safety (OS) at the user companies other than inducting the workers to the best of their ability. The workers felt that much of the responsibility was transferred to them, and they were kept in the dark about or avoided information about OH and OS. Overall, the agency workers were in a weaker position than permanent workers concerning OH and OS.

Poor induction, training, and supervision
Based on national laws, all new workers must be given a proper induction to work, and this should reach those working through agencies. In most cases, the induction was covered by written material, a so-called induction folder that the agency gave the worker to read. As a following stage, a workplace-specific induction was expected to be provided by the user company. In principle, this arrangement was considered good by the interviewees. However, challenges were experienced in cases where the worker simply had not paid attention to the induction material given or showed no interest in listening. Such gaps in induction were assumed to be covered by assigning the worker a contact person at the agency to answer questions later if the need arose. This was also a challenge identified by the workers, as the situation might be confusing with two supervisors: one from the user company and one from the agency. The agency workers felt that in some cases they did not necessarily know which matters belonged to which supervisor. They also pointed out issues related to continuously changing workplaces, where they had to familiarize themselves again and again with new working methods and workplace cultures. They considered this stressful and as influencing their work performance.
Concerning training not related to induction, the agencies offered additional training to their workers to increase their chances of getting a job. This was considered not only an act of philanthropy but also an act that made sense from an economic perspective, as the agency made money every time a worker was hired. From this perspective, the economic and social elements of CSR can be considered as benefitting the various actors hand in hand. Occupational safety training and fire safety training, as well as some specific courses to gain minor licences, were mentioned as examples of such training. In addition, user restaurants offered some minor professional training, such as hygiene and alcohol training. Agency Expert 3 complied well with the shared understanding of the benefits concerning induction and training but also emphasized some issues that might be confronted: We have clients [user companies] that have really good worker induction models and practices that handle matters of occupational safety. We also have our own model. Whenever a new worker starts, the orientation should be done, either by us or the client. Many clients [user companies] do not have a formal induction process. We can ask and instruct the client to induct the worker, but we cannot command the client to do it. If it seems that the worker cannot work safely, then we inform the worker that it is not possible to continue working there until the client has taken care of safe working.

Obstacles for employee organization
The attitudes of workers employed by the user companies could be quite hostile or indifferent towards agency workers, who might even be thought as a burden or intruder. Particularly in cases of poor working atmosphere, agency workers often became the targets of ill feelings vented by other workers. Agency workers thought that the prospects of influencing their own work were limited and that whatever they said did not affect anything. Moreover, they feared that intervening in grievances might negatively affect their chance of getting work in the future. Because of their precarious situation and fear of losing work, they stayed silent even when confronting challenges and obstacles at work. Agency Expert 2 spoke about the concerns of workers: There are workers that have been working for a long time for a user company but do not get the benefits that other workers of the company get. They do not dare to speak about it because they feel that they are on eternal probation, and that is what they are on, for it is easy to get rid of them. The user company only needs to notifiy them about it.

Poor or fragmented communication
Both the agencies and the user companies highlighted that they had their own methods and lines of communication. Challenges were identified as arising when they needed to communicate beyound the borders of their own organizations. Much of the interorganizational information was considered as being lost on such occasions. All the actors wished for a joint system of communication that all could use. Then information would be easily accessible and the same for all actors (and in multiple languages for foreign workers). Correcting possible errors and taking responsibility were impossible without timely and accurate information, as was highlighted in an interview with Agency Expert 4: There have been a lot of misunderstandings. I have faced many situations where the person from the user company has not known what the worker should do and the worker did not know even that much. Then I just needed to figure it out while every-one was waiting. Well I knew. I had papers telling everything; others did not. Somehow others did not get the message.

Gaps in regulatory coverage
A clear challenge in regulatory coverage was highlighted by the workers. This related to situations in which the agency and the user company are owned by the same entity or chain. The owner is responsible for both companies, the activities between these companies, the payment of wages, and other legislation related to the worker doing the actual work. In practice, however, this forms a kind of grey area in the regulations concerning the companies involved. It brings challenges regarding the interpretation of the collective agreements. Grey area agencies were considered as making the interpretation that if the employee had been absent for a while from a previous work shift, they could view the employment relationship as starting again from the beginning. One practical way of repeatedly starting employment was identified in a situation in which there were two separate agencies and a user company under the same parent company, with the agencies taking it in turn to pay the salaries of the workers. In this joint ownership of the agencies and the user company, the employers were in a decisive position in all aspects of work, without the need to listen to workers. If workers did not comply with this practice, the work shifts could be removed, but the workers were not dismissed. Legally new workers could not be hired in the place of dismissed without a very good justification, so workers just lingered on the list without getting workshifts and completely new worker was hired. Workers who had experienced and seen this situation repeatedly in the restaurant field, as well as the labour union, thought that this whole arrangement only existed to avoid CSR, legal obligations, and workers' rights. Agency Worker 2 presented the following view on simultaneous ownership: I was wondering why I got my pay cheque from a different payer every other month. It seemed like two agencies were taking turns in paying. I did not know that the restaurant I was working for owned them both. At the time, I never knew why this arrangement was used.. .. It became clear when I needed to use my holidays; it turned out I did not have any.

Poor knowledge and disputes regarding legal rights and obligations
Disputes occurred most due to the challenges caused by each actor having their own interpretations of regulations and situations. It was perceived to be common for each side to try to get results that best suited their own preferences. For instance, it was raised that young workers did not necessarily know their rights and obligations, and even if they did know, they feared that upholding them would lead to their employment being effectively terminated. Lately, agencies have begun to instruct and pressure user companies to comply with the law and follow matters relating to CSR. Agency Expert 5 had come across a situation where poor knowledge was sought by a user company: Some user companies prefer to use agencies and workers that do not know their rights and obligations well, because they do not necessarily know that they should be paid extra compensation for hours after a certain time or during holidays. Unfortunately, there still exist actors that try to circumvent the rules.

Poor regulatory oversight
The experts brought forward challenges in dealing with the government. The agencies and user companies expressed how the government bureaus involved often fail to work together when dealing with TAW matters, resulting in overlapping and even contradictory views. Another challenge identified by the agencies and user companies was government inactivity in addressing arising challenges. This was perceived to open opportunities for irresponsible actors to enter and operate in the field. Also, agencies and user companies considered legislation to be lagging and not covering all aspects of TAW. In practice, much of the regulatory oversight was carried out in user companies and agencies, so the importance of CSR cannot be understated. Agency Expert 4 shared their opinion about the poor regulatory oversight of agencies: Starting this business is very easy; therefore, "special businessmen" started doing business and make errors. Those kinds of businessmen do not want to tell their customers proper modes of operation, if they even know them, and allow the client to do outright illegalities. For example, workers will not get paid for overtime. and they claim that workers have agreed. Better certification systems must be developed to keep those "special businessmen" out of the agency business.

Summary of the findings
As a summary of the above findings, a synthesis of the challenges and their relationships with TAW actors is presented in Fig. 1. Challenges that reciprocally affect all actors are placed at the centre of the figure. Poor knowledge of rights can be associated with unilateral agreements between actors and mainly affect the workers. Poor governmental control, which was highlighted by all actors, affects the formation of such unilateral agreements. In some cases, agencies can dictate the content of the agreement, thus making it unilateral. Poor governmental control, or a lack of such control, enables the abuse of worker pools, yet it must be noted that these can also be used in responsible ways. Current governmental guidance also allows for the utilization of joint ownership arrangements between agencies and user companies, which mainly affects the workers. Unequally organized occupational safety and health (OSH) arrangements affect agencies, user companies, work communities in restaurants, and ultimately temporary agency workers.

Discussion
We analysed CSR in the restaurant industry by identifying potential challenges, areas for improvement, and strengths from the perspectives of the different actors in the triangular relationship between workers, agencies, and user companies. The results showed the diversity of TAW in restaurant work, with many kinds of actors operating in this field. There was evidence on ethical dilemmas in practice. Interviews revealed how the regulations concerning TAW were not always fully known or they were even fully ignored. Concerning such irresponsible behaviour, the worker was often left in a vulnerable position. These findings contribute to the discussion on the paradigms of CSR and CSIR that urge to seek better understanding on CSIR control and prevention mechanisms (e.g., Riera & Iborra 2017). Irresponsible behaviour might relate to agency workers' own actions, but they may also relate to the possibilities that the triangular and joint ownership employer arrangement provides. Concerning our first research question on the realization of CSR in TAW as examined through the PDR-model perspective, we conclude that user companies and agencies seem to be overwhelmed by economic incentives. This is not surprising considering the profitdriven and competitive nature of the restaurant and TAW business. However, we point out that the social aspects should not be forgotten, as restaurant work is still highly labour and socially intensive despite technological advancements in this area. Social matters stood out from the material, especially regarding the implementation of OSH and in work arrangements. Concerning OSH, issues such as awareness of responsibilities and communication matters were highlighted, whereas work arrangements were critically discussed based on the complex nature of flexibility, challenges in entering work communities, and uncertainties from the regulatory point of view. In general, functional, and organizational aspects of TAW should be given more attention. For instance, collaboration, communication and trust between different actors and stakeholders should be improved. Unfortunately, the interviews revealed that there were actors in agencies and user companies that did not pursue good CSR, while also attempting to circumvent regulations for their own gain. This distorts and intensifies competition and leads to pressure that makes it harder for responsible agencies and responsible user companies to maintain their responsible conduct.
As an answer to our second research question, we identified several challenges and points for development in TAW, but we were also able to capture various opportunities that the actors should hold on to and develop further. Table 1 puts together these findings and proposes actions for future research activities in this field.
In relation to the previous research, our results supplement Forde and Slater's (2016) comprehensive review findings. Our findings introduce partnerships with agencies and user companies together with their constantly more complicated contract arrangements. Worker pools take the development of partnerships even further. In the light of the views of the agencies, the goal of partnerships and worker pools is to tie the user company ever tighter to the agency to ensure profits and the continuity of business relationships. A worrying aspect is whether this development is also intended to weaken the position of the employee, which  Knox (2018) saw new emerging ways of developing elaborate strategies to avoid regulation in TAW as a longer trend, and the joint ownership that was present in our data is probably not the last to appear. Regarding occupational health and safety in a TAW context, Strauss-Raats (2019) found only little evidence that any TAW party was unfamiliar with their obligations, but the pursuit of economic interest overruled the application of this knowledge in practice. A similar effect could be found in our study in a CSR context, as economic benefit seems to be the main reason for not complying with the principles of CSR, and it overrules social responsibility aspects. At least partly this might relate to inadequate legislative protection for the agency workers. Based on our study, it seems evident that the legislation is lagging and does not cover all essential aspects of TAW, thus leaving room for actors that may be trying to benefit of this grey area. We point out that this is not only a Finnish phenomenon, but researchers abroad have constantly reported on similar challenges with very little evidence on any progress (e.g., Howes 2011;Oyetunde et al., 2022;Sartori, 2016;Strauss-Raats, 2019).

Strengths and limitations.
The application of the CWEQ-II in the form of interviews instead of a questionnaire was a strength. The interviewees were able to add aspects they considered important. In this way, more information was obtained without the limitations of the questions. A second strength is that the whole research team cross-checked the results for more reliability. A third strength relates to the national level project in which this study was a part of. To benefit from the project, means for sample selection, data collection and analysis were agreed with the project group, consisting of researchers from four universities. In addition, the large national steering group with members from employee and employer unions and authorities was consulted while designing the research activities. A certain limitation of this study relates to locality, as the research was conducted in only one country, Finland. However, the interviews did not focus on a particular geographical area, with the interviewees representing actors across Finland. Another limitation relates to the sample size. To a certain extent the size was limited to practical matters. However, the authors considered that a certain saturation point was likely achieved with 29 interviewees representing the essential parties of TAW. The themes (TAW, restaurant work, and CSR) of this study are common concepts, shared globally. It is important to acknowledge that Finland and other Nordic countries are paradigmatic welfare states, where employability of workers is taken care of (Townsley & Stohl, 2003). The Nordics also have good welfare systems and are more egalitarian and inclusive, with decent wages and strong labour unions (Rasmussen et al., 2019). Thus, TAW is not completely unregulated. Therefore, our supposition is that these counties have less contingent and precarious labour markets than other countries. Still concrete evidence was found on the challenges from the CSR perspective. The nature of these factors should be acknowledged when considering transferring results to another context.

Conclusions
Responsible companies understand the importance of a good employer image, and most of the agencies and user companies followed this principle. Still irresponsible actions were also identified in this study. A major challenge from the CSR point of view in restaurants is how financial gain and employee well-being, including those of temporary agency workers, can be tied together. Is this even possible under current regulations? Many of the challenges identified in this study seemed mutual for all the actors; however, when examined more in depth, the worker was usually the one who was in the weakest position.
The challenges are wide and multidimensional, and there are many determinants and perspectives, including ethical considerations that should be acknowledged when resolving them. A more serious approach to CSR is needed by the actors, especially by agencies and user companies, to rectify the complexity of the challenges related to TAW in restaurant work. At a workplace level, agencies and user companies also need to develop their HR practices including OSH, preferably together with workers, to enable sustainable TAW. Lastly, more proactive role of government is needed to correct shortcomings in legislation.
Funding This study was funded by the European Social Fund (Project number S21965).

Open Access funding provided by University of Oulu including Oulu University Hospital.
Declarations The principles of research ethics were followed based on the Finnish National Board on Research Integrity and the General Data Protection Regulation (GDPR; EU, 2016/679), and the project of which this study is part was approved in Fall 2020 by the Human Sciences Ethics Committee of the Helsinki Region Universities of Applied Sciences.

Conflict of interest The authors declare no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.