Reinventing the Division of Labour? How Organisational and Micro-Level Factors Influence Work in Local Public Welfare Organisations

2.1 Vertical and Horizontal Integration of Work in the Public Sector

On a general level, several changes in public welfare organisations, related to rationalisations and organisational efficiency reforms (Alvehus and Andersson 2018; Hugman 2020), have in different ways reduced the division of labour in these public organisations. These observations are clearly opposed the idea of processes where occupations in public organisations “hive-off” tasks (Hughes 1958). However, there are virtually no concepts available to describe the trend towards a reduced division of labour in public organisations. In the literature, concepts such as “de-differentiation”, “de-specialisation” (McRobbie 2002) or “de-professionalisation” (Hales 2006: 100; Hugman 2020) appear, but as will become clear later on, these concepts are too vague to be useful in this study. I therefore follow Boglind (2019) and utilise the concepts of horizontal and vertical integration, which allow for a more accurate analysis of these processes.

Horizontal integration reduces the degree of specialisation in individual work tasks, where the responsibility for these tasks is instead transferred to a multi-skilled group or team (Hales 2006: 102). Horizontal integration therefore blurs the boundaries between production and maintenance work (cf. Strauss 1985). For example, greater flexibility can entail expecting employees to clean their own workspaces and replace broken light bulbs, tasks previously performed by other categories of staff, such as assistants and secretaries.

Vertical integration consists of the transfer of management functions, such as planning, coordination, and supervision, to ordinary employees in the organisation (Hales 2006: 102; Hugman 2020). A consequence of this is that employees gain flexibility and more opportunities to plan and monitor their own work, and some literature discusses this change as “empowerment” at work (Harley 1999; Kanter 1993).

The horizontal and vertical integration of work can also be associated with the concept of separability, where low separability is characterised by “diffused, overlapping, intertwined roles” (Lawler, Thye, and Yoon 2009: 58). A high degree of separability is instead distinguished as “discrete, specialized, independent roles” (Lawler, Thye, and Yoon 2009: 58), in other words, what we might call a classical, strict division of labour.

2.2 Perspectives on the Division of Labour – Market, Bureaucracy and Occupation

Early on, Freidson (1976: 306) argued that the abstract and structural definitions of the division of labour presented by Smith, Durkheim, and partly Marx, were weakly related to empirical reality, since they lack a “reference to a concrete, historical process, bounded in time and space”. The classic perspectives on the division of labour provide little information on how actual relationships are affected by the division or sharing of collective work (Chin 2011: 16), and more contemporary studies tend to focus primarily on structural features (Bach, Kessler, and Heron 2007). Therefore, questions about the social aspects of the division of labour, such as cohesion, collective participation, and social identity, have often been neglected (Bach, Kessler, and Heron 2007). One explanation for this is that much of the research concerning the division of labour in organisations has started from a “top-down” management perspective. To counter the somewhat one-dimensional perspective on how the division of labour is organised and controlled, Freidson (2001) proposed a more systematic model based on three ideal types: the market, the bureaucratic organisation, and the occupations (see Freidson [1976] for the initial typology).

First, the division of labour can be organised via free competition in the market. Following Adam Smith’s ideas of a free market, workers are free to interact and compete with each other and perform any kind of work without constraints (Freidson 2001). Jobs of this ideal type are characterised by fluidity, a high degree of specialisation, a short lifespan, a low degree of job standardisation and a low demand for complex skills. Freidson (2001) argues that this ideal type is difficult to imagine in complex societies.

The second ideal type includes the rational bureaucratic organisation as a social force in the organisation of work. Relating to Weber, Freidson argues that the essence of the bureaucratic model and its hierarchical rational-legal authority is to “formulate, distribute and supervise specialised tasks” (Freidson 2001). Organisations “plan in formal and rational detail what each task shall be and who shall be qualified to perform it” (Freidson 1976: 309). Thus, work is organised to suit the needs of the organisation, and while there is room for variation, “neither consumers nor workers control such change” (Freidson 2001). Processes leading to horizontal and vertical integration are, for example, closely related to the role of organisational authority.

Compared to the market model, the number of jobs in the bureaucratic model would be fewer, less fluid and more stable. Furthermore, bureaucratic organisations tend to increase their differentiation when they increase in size, which in turn creates the space for a plurality of different formal tasks. Fragmented tasks also entail the possible establishment of asymmetric power relationships in the organisation, where administrative positions, for example, are given more decision-making powers than other positions (Glucksmann 2009: 881; Rueschemeyer 1986).

The third model, the occupationally controlled division of labour, is orientated towards the role of occupational, or worker, control- and workplace interaction in determining how tasks are designed, who should perform them, and how they shall be performed (Freidson 1976: 310, 2001). This perspective suggests that employees are involved in designing the conditions for, and content of, the work, regardless of what formal organisational efforts are made to justify, control, or understand their activities (Freidson 1976: 310). In a study by Chin (2011: 17; cf Lawler 2002), this occupational perspective is seen as a “relational micro-order conception of the division of labour”, and Chin’s aim is to understand how collective duties are shared between group members.

Freidson (2001) argues that an occupationally controlled division of labour is an essential part of professionalism. This so-called third logic involves a great deal of negotiation around occupational jurisdictions. Some negotiations concern particular tasks; others involve task priority. The occupationally controlled division of labour is characterised by a small number of stable specialisations, although relatively differentiated within occupations, which also leaves room for internal struggle and negotiation (Freidson 2001).

The point, for both Freidson and Chin, in using the occupational model is to place the division of labour in a practical and socially regulated context. The division of labour in the occupational model is thus seen as a continuous process, where individuals and groups are engaged in “conspiracy, evasion, negotiation and conflict” when it comes to dealing with the diverse conditions and situations that exist in the workplace (Freidson 1976: 310).

A more recent attempt to conceptualise occupations and professions is the framework suggested by Anteby, Chan, and DiBenigno (2016), based on three different lenses. These lenses capture a temporal ordering of the research in occupations and professions, where the “becoming” lens focuses on how individuals learn to be part of a collective, the “doing” lens focuses on the activities actors are engaged in, and the “relating” lens targets how occupational members relate to others outside their group (see, for example, Strauss [1985] on actors working in proximity or distance from each other). The relating approach clarifies how collaboration between occupational groups can occur despite occupational differences (Anteby, Chan, and DiBenigno 2016: 215). Freidson’s perspective is mostly related to the “doing” lens, and one of its limitations is the emphasis on conflictual interaction and competitive dynamics (Anteby, Chan, and DiBenigno 2016). Introducing the “relating” lens “might reveal not only competitive dynamics but also collaborative dynamics between professions in how they enact their professional logics” (Anteby, Chan, and DiBenigno 2016: 224; cf. Torfing, Sørensen, and Røiseland 2019), a perspective relevant in the present study of service assistants in local public welfare organisations.

The “relating” lens can also be linked to Huising’s (2015) findings of how professional groups interact with clients and other occupational groups. In the classic literature, occupational groups regularly hive-off tasks considered “dirty work” (Hughes 1958) to other occupations in order to control the division of labour and reinforce their professional image, as mentioned in the introduction. However, according to Huising (2015), while doing dirty work or “scut work” may damage one’s professional image, it may also increase autonomy, self-regulation, and the protection of professional jurisdiction in relation to organisational demands and other occupational groups. Professional groups hiving-off tasks to influence the division of labour in their favour can deprive them of important resources, such as reinforced expert knowledge and skills, instead of facilitating relational authority (Huising 2015).

According to Freidson, his model makes sense of different ways to organise and control the division of labour in society. Obviously, Freidson points out that reality is a variable mix of the three co-existing and competing logics. For example, workplace practice can often create ambiguity and counteract the formal plans and objectives of the rational organisation (Freidson 1976: 310). At the same time, scholars have observed that occupational and professional autonomy have been subsumed under market and bureaucratic logics (Alvehus and Andersson 2018). Alvehus and Andersson (2018) argue, drawing on institutional theory, that modern welfare professions are in a situation of intertwined and entangled institutional logics, where both organisations and professions try to increase their influence on the division of labour. However, Alvehus and Andersson (2018) also acknowledge that “professionals repeatedly show a strong resilience toward changes in work practices”. This perspective raises awareness of the simultaneous effects of professional and organisational logics in the organisation of work; both may exist in the same organisational setting, but with different consequences for everyday work.

2.3 Studying the Reversibility of Division of Labour

To explore the reversibility of the division of labour within a chosen set of public sector organisations, the analytical starting point is placed at the intersection between the bureaucratic organisation and the occupationally controlled division of labour (Freidson 2001), leaving out the market ideal type since in Freidson’s model it does not add much analytical weight. These two perspectives have been discussed as two different sets of logics influencing the division of labour in organisations (Alvehus and Andersson 2018). The empirical objects of the study, the local government labour market initiatives, have all been created by the local government employing a top-down organisational perspective of the division of labour.

By analysing cases where the division of labour has been subject to reform, the intention is to improve our understanding of organisational work-related reforms in professional and occupational settings and how the division of labour operates in local public organisations. At its core, this question also concerns the exercise of power between, on the one hand, the local government initiating the reform attempt, and on the other the occupations within the different public organisations. The possibility of influencing the division of labour thus mirrors the dynamics of this relationship, as well as the competitive and collaborative dynamics of the occupations.

4.1 Patterns of Reversed Division of Labour

The nursing unit in the study performs health care in assisted living facilities for patients that have been allowed such support by the local government. Work at the nursing unit in the medium-sized municipality is characterised by a heavy workload, which according to the staff is due to an increased burden of care, an increased amount of administration and documentation, and greater demands on coordination and cooperation between multiple actors, among other factors. The manager describes one factor why the burden of care has increased:

There has been a discreet shift in the healthcare system. Today, patients are sent home from the hospital much earlier. Today, the patients are really, really ill, which requires much more from the nurses, while the number of nurses has not increased, they have to work much more.

Employing service assistants has meant that the nurses have been relieved of tasks that are not traditionally part of their duties, such as re-stocking storerooms, emptying the garbage, handling supply and medicine, and laboratory deliveries, as well as transport services between different assisted living facilities. A nurse at the unit explains her thoughts on these tasks:

I understood that I previously did a lot that was completely unnecessary based on my nursing skills. I can question why I studied nursing at all considering what tasks I performed.

Defining tasks to delegate to the service assistant was, however, not effortless since nurses with variable experience differ in their perception of which tasks should be included in the occupation, and which should not. Nurses with shorter occupational experience tend to include more tasks that could be labelled “dirty work”.

After employing the service, assistant nurses explain that they have more time for patients. A nurse recounts:

I don’t dare imagine what it would be like without the service assistant now when you’ve gotten used to and experienced the tremendous relief it provides, and how much time it frees up for patient work. Actually, I have a hard time understanding how we managed our days before he was hired.

The study of the nursing unit therefore shows that the local government efforts to increase the division of labour has been realised through both organisational and occupational authority. One observation is that the nature of the work matters, since it is performed by nurses who have an occupation with relatively clear boundaries between the healthcare mandate and other additional duties. This means that there are formal and practical obstacles for a service assistant without healthcare training being integrated into the regular work of caring for patients. The service assistant position does not require nursing qualifications, and it has therefore been possible to transfer nurses’ secondary tasks to the service assistant position, in line with the objectives of the local government initiative. Another empirical observation is that the work of the service assistant was performed at a distance from the daily work of the nurses since a major part of the assistant’s tasks involve transport of various kinds, which in turn promoted the division of labour. A nurse describes the tasks:

The service assistant drives [medical] samples all over the municipality to different health centers […], he collects all the medicines from the pharmacy […], then the service assistant takes care of all the tasks regarding our cars, tire changes, taking the cars to service and refuels them and yes, all that sort of things. […] He also drives certain medical waste we are not allowed to empty in our bins to the recycling centre, as well as patient mattresses all around, yes and collects various signature lists that are to be sent to nursing staff in different premises.

A third observation is the relative ease with which tasks that do not require healthcare skills, that is, work tasks with high separability, are defined. This is related to the fact that these tasks were added through former organisational reforms, for example, the increased need for coordination among a greater number of providers, transport, and logistics. Other work tasks that have been added through horizontal integration, resembling caretaker duties, can be related to reforms aimed at streamlining operations.

The reasons for the clear division of labour at a primary school in the metropolitan municipality and a ward in an assisted living facility for the elderly in the medium-sized municipality, are largely based on factors resembling the case of the nursing unit. In the primary school, the teachers have experienced an increasing workload, especially since the introduction of the most recent curriculum in 2011 (Skolverket 2019). The assistant principal in one of the schools emphasised new administrative demands as one explanation for the increasing workload:

The documentation requirements have increased. We must document a lot, and it is very demanding for them [the teachers]. […] This is what caused the heavy workload for the teaching staff. And all this with IUP [individual education plans], progress plans and so on, which you also need to fill out before every meeting with parents.

The increased workload is mostly administrative, including documentation, but also related to social tasks not intimately associated with teaching. As observed in the study, the service assistants have reduced the social workload of the teachers. Teachers have been able to focus on “developing the knowledge of the children” instead of spending a significant amount of time maintaining peace and quiet in the classroom, according to the school’s assistant principal. One difference from the nursing unit is that service assistants have been working closer to the teachers. However, the service assistants did not have any type of teacher training or educational work experience, and they were mainly performing social tasks. According to the assistant principal, the service assistant was not allowed to have a teaching assignment, and these circumstances promoted the introduction of a clear division of labour at the school:

The service assistant cannot not teach. It is very clear from the union’s side. But in a school you need many hands, that’s how it is. There are many children and they are on the move. What we really needed is support during breaks and then extra support in the classroom and someone who is able to calm down the kids in corridors and things like that.

The service assistant position at the assisted living facility for the elderly includes many everyday routine elements of the workplace, such as handling food and cleaning, which differs significantly from the patient centred care work of the nursing aides. The nature of the work and the need for healthcare skills here play a major role for the possibility of finding clearly defined tasks that can be transferred to the service assistant. The required nursing aide qualification thus reinforces the boundary between service assistants and the nursing aides. However, the profession is weaker at the assisted living facility, while the service assistant also works closer with the nursing aides than was the case at the nursing unit.

An additional aspect, related to occupational control, which created conditions for an increased division of labour in the studied labour market initiatives, is linked to the service assistants’ language skills. A preschool director develops her views on the importance of language:

Because the language is the key to being allowed to participate a little bit more. […] I think, like, you come here to be a service assistant in order to practise the language.

According to the director, weak language skills serve as a barrier to further horizontal integration, especially in residential areas where languages other than Swedish are common. At three of the studied preschools where the service assistants’ work is clearly separated from the childcare assistants’ and the preschool teachers’, the division of labour is largely based on the managers’ assessment of the weaker linguistic abilities of the service assistants.

4.2 Patterns of Service Assistant Integration

The empirical data show that a majority of the organisations in the study encountered difficulties in implementing an increased division of labour. The recruitment process of the service assistants, as well as the possibilities of defining and delimiting different occupational core mandates, have played a role in the implementation problems.

Regarding the hiring of service assistants, employers, who often share the same occupation as the ordinary employees in the respective organisation, were given the opportunity to interview possible candidates from a group deemed appropriate by the municipal labour market organisations. The study has shown that these recruitment processes, unlike many other labour market programmes where applicants most often are placed, have given employers the opportunity to acquire staff they perceive to be both competent and suitable for the work of a service assistant. A general finding in the study is that the more employers and regular staff assess the service assistants as qualified, the less inclined they are to establish a clear division of labour in the workplace. In addition, this assessment has been ongoing over time, which means that service assistants in many cases were assigned more complex tasks after a period of employment. This is particularly evident among the small group of service assistants with a teaching background, but also among service assistants in preschools. A teacher describes working with the service assistant in the classroom as follows:

Yes, but she does what I do actually, even if maybe she isn’t in charge of the lesson and doesn’t plan it. […] And once [the children] start working you see no difference between us. Then we are equal. […] So she does what I am doing, but maybe on a smaller scale.

Staff and managers perceive it as a waste to only assign practical tasks to service assistants who are deemed qualified and suitable. A preschool director talks about the work a service assistant performs:

She is completely with the group of children, yes. […] And also participates in our workplace meetings and the team planning […]. But as I said, we noticed very quickly that we had an ambitious girl here. It would have been a waste to have her make sandwiches and push food trolleys all the time.

The qualification level of the service assistants can be a matter partly of formal educational background, such as in the schools, and partly one of personal suitability, as in the preschools. The study therefore shows, somewhat surprisingly, that despite a relatively strong profession within the school, the teachers are willing to involve service assistants in the core mission. One likely factor in this context is that the teachers consider the service assistants as peers, given their teacher education background, and want to support their development towards gaining qualified teacher status certificates. The weak division of labour between educators and service assistants in preschools can be related to the fact that the main educational mission is difficult to separate from other types of work tasks. In practice, this also means that the two regular occupations in preschools, preschool teachers and childcare assistants, have very similar work tasks.

The formal educational requirements in Sweden are also not as clear in preschools as in primary schools, which means that the personal suitability of the service assistant is rewarded instead. These circumstances made it harder to distinguish and define clear-cut tasks for service assistants in preschools. A preschool manager explains her views on personal suitability:

But being with the children, it’s probably more about your personality. Whether you are suitable to be with children or not. It doesn’t matter, you can have studied for four years but you can’t interact with a child in the right way.

Another observation in the study is the co-existing and competing organisational and occupational logics, which has limited the prospect of increasing the division of labour. In the empirical material, the municipalities have tried to direct the focus of the service assistants’ work based on the organisation’s need for workload reduction through the division of labour. Documents such as the centrally designed job descriptions for the new positions have therefore been important. However, the implementation of these job descriptions has proved difficult in practice. The study shows that the service assistant job descriptions were developed in informal negotiations at the workplace, a process involving local occupations such as managers, staff, and service assistants in preschools, schools, and assisted living facilities. A principal at a primary school explains that it is mainly the teachers working with the service assistants who have had the greatest influence over the service assistants’ work:

I haven’t done that [directed the everyday work for the service assistants]. I have directed where they should be and I have moved them, when needed, to other places. But what they do inside the classroom with the main teacher, I think the two of them have to decide together.

Further, a preschool manager also describes how tasks were assigned to the service assistants:

In the beginning, I had no idea how we could use a service assistant. And the staff didn’t either, so together we have … well, without us actually writing anything down, it’s turned out the way it’s turned out. We have been reasoning, they have asked me questions, ‘How should we think, what should we do?’, ‘Yes, but have you asked the service assistant? What does she think?’

The occupations have thus had the scope to define the work content of the service assistants, but this process has largely occurred in collaboration with the service assistants.

Clearly, the staff’s assessment of their need for workload reduction has also affected how the service assistants have been utilised. In areas where workload reduction is perceived to be most necessary in connection to the core assignment, for example within primary schools and preschools, this limits the division of labour to a higher degree. In areas where secondary tasks are perceived as creating a heavier workload, for example, at the nursing unit, the possibilities of increasing the division of labour are greater. Thus, the nature of the work is mirrored by these needs and is also dependent on the staff’s discretion. In practice, this has meant that other considerations have been taken into account, which are more related to staff attitudes and the dynamics of the workplace, rather than only focussing on the municipal organisations ambition to increase the division of labour. This has meant that staff, particularly in preschools and schools, have due to solidarity involved the service assistants in more, as well as more qualified, work tasks.

Staff at several preschools in the study were in many cases uncomfortable with only allowing service assistants to perform the more routine tasks unconnected to the group of children. It is also clear that they want to work together on an equal footing, and not make the service assistants feel inferior. Some educators explain:

It feels wrong that [the service assistant] just gets to do the grunt work. I wouldn’t want to do it if it was me [who was the service assistant].

I think it’s better for her if she [the service assistant] gets to be one of us, rather than that we should give her tasks all the time. She knows what to do, sometimes she does things herself. She knows that it will be snack time soon and starts laying the table.

Additionally, a preschool manager describes the staff’s thoughts on creating a more distinct division of labour at their preschool:

And we’ve talked about how to do with the meals, if she [the service assistant] was going to do the setting and cleaning of the tables in all the departments. No, but then everyone in the staff had second thoughts and just said ‘No, but that’s not what we want her [the service assistant] for, we can set and clean the tables ourselves’.

One of the explanations for this observed solidarity can be related to the distance/proximity factor, discussed earlier in relation to the nursing unit, since educators and service assistants work closely with each other. Delegating unskilled tasks are apparently more challenging when educators work in close proximity to the service assistants.

Moreover, there are indications that some of the more routine and unskilled work tasks have an intrinsic value in the studied organisations. “Boring” and repetitive tasks, referred to as “dirty work” in the literature, are perceived as a break from the main work, and staff do not always want to delegate them to other occupations. One educator explains:

At the same time, it can be nice to get away from the group of children. You can have a breather if you’re working in the kitchen.

According to the study, staff in preschools and schools share both the main and routine work tasks to a large extent. However, there are obviously tasks that the service assistants do not perform, such as systematic quality assurance, operational planning, and formal contact with parents. In comparison with the service assistants’ distanced work at the nursing unit, the preschool and primary school staff work closely with the service assistants in everyday work. Altogether, these factors of the occupationally controlled division of labour limit the opportunities to increase the division of labour in the municipal organisations.

4.3 Analysing the Reversibility of the Division of Labour Initiatives

In this study of attempts to reverse the existing division of labour in a set of local public welfare organisations, six factors have been observed that are able to influence the reversibility of the division of labour. A first factor that was found relevant in influencing this reversibility was the character of the boundaries between core and secondary tasks in the existing occupations (Strauss 1985). Some domains are organised in a way that enables reversing the division of labour through the delegation of tasks from one occupational category to another, under certain conditions. Educational requirements are fundamental to this factor, as a stronger profession with clear educational requirements creates opportunities to define the boundary more clearly between core and secondary tasks, and thus also find tasks to hive-off that fall outside of the main scope (Hughes 1958). More specifically, it is the professional ability to precisely define and delimit the core mission, which promotes a distinct division of labour (Chin 2011). According to Lawler, Thye, and Yoon (2009), it is thus possible to talk about the possibility of establishing a high degree of separability when it comes to the main assignments of a position. However, the initial condition in the case of the nurses clearly indicates that processes of horizontal integration had blurred the boundaries between care and maintenance work, following the observations made by Hales (2006). An occupational category where the work tasks are extensive, well-integrated, and long established tends to hamper the delegation of tasks to the service assistants, hence the low degree of separability. Some sectors and occupations have for decades had a limited division of labour, as seen in the example of the preschool. Other professional categories, such as the employees at the nursing unit, may have a faint “collective memory” of reforms and reorganisation that affected the content of the work. In this case, it may be easier to re-separate these tasks from the core mandate.

Furthermore, the arrangements for negotiating the form of the service assistant positions also had a major impact on whether and how the division of labour was realised. In all cases, descriptions of the different service assistant positions were developed centrally within the bureaucratic initiative, which were then communicated to the local managers. The implementation of these job descriptions, and thus the ability of management to practically define the tasks that can be transferred from the regular staff, differs between workplaces. Especially in the educational domain, the tendency is that occupational control and workplace practice influence service assistant positions, rather than abstract descriptions made by local government managers. In this study, no observable negotiation of occupational jurisdictions was present in relation to the local government organisation. However, the process of establishing a more distinct division of labour in the municipal welfare organisations indicate the presence of co-existing (Freidson 2001) and intertwined logics (Alvehus and Andersson 2018). The staff’s discretion has been central, and they have largely shown an inclusive approach towards the service assistants. Such a scope for decentralised decision making regarding the service assistant position did counteract an increased division of labour in the majority of cases. The second factor that impacted the reversibility of the division of labour is therefore related to the level of decision making in the implementation process of the new positions.

Thirdly, the assessed qualification level of the service assistants seems to be of importance for the division of labour in the study. Qualification assessment, performed locally by the regular staff and their managers, includes the evaluation of education and personal suitability, and is an assessment related to the performance of service assistants in the day-to-day work as well as in the recruiting process. The recruiting process in the two studied initiatives, initiated at the central municipal level, resemble those of the regular labour market, which tend to create greater opportunities in terms of choice for local employers. This, in turn, has likely led to applicants being hired who were perceived as most qualified for the service assistant positions. Service assistants whose education resembles that of the regular staff, as is the case in the primary school, tend to be assigned work tasks similar to the work of regular staff. However, the constant assessment is also important for the regular staff’s willingness to delegate work tasks.

When it comes to occupational categories that have less education and a weaker professional position, for instance, the limited division of labour seen in the preschool example, the importance of service assistant education is reduced, and the focus instead tends to shift to valuing personal suitability higher as a desired qualification. Service assistants assessed as suitable also tend to be invited to perform the regular staff’s tasks. Workplace occupational qualification assessments following the logics of education or personal suitability therefore tend to counteract the intention to establish a strict division of labour between service assistants and regular staff, whereas a lower degree of assessed qualification of service assistants tend to, vice versa, create opportunities for a stricter division of labour. One part of the occupational assessment process is also how the regular staff and their managers perceive the language skills of service assistants. Weak Swedish language skills tend to make it easier for managers or the regular staff to create positions that are separate from the regular staff, while acceptable or good skills imply the opposite.

The study furthermore shows that a fourth factor related to the regular staff’s sense of solidarity limits what work tasks they would consider transferring to service assistants. The regular staff are keen that the service assistants should perform meaningful tasks that are clearly related to the main work, even though regular staff, at a first glance, would seem to benefit from avoiding the less skilled tasks. Related to the solidarity aspect is the physical distance factor (Strauss 1985) between service assistants and regular staff in the workplace. Close interactions between service assistants and regular staff in the workplace, for example, in the preschools, tends to create less opportunity for a strict division of labour, and the opposite for interactions at greater distance, such as in the nurse unit. This factor is also related to the strength of the professions in the workplace, and how clear the boundaries are between the core and secondary tasks. The solidarity factor and examples of closeness between service assistants and regular staff emphasise collaboration and the relevance of a relating approach (Anteby, Chan, and DiBenigno 2016), as not much conflictual interaction has been observed (Freidson 2001: 1976).

The sixth factor highlights the observation that some repetitive tasks are perceived to have an intrinsic value among the regular staff. This observation, particularly in preschools, indicates that monotonous and routine tasks fill a need for the staff since they might mean a break from an otherwise high-pressure working environment. One of the consequences of this fact is therefore that the regular staff might limit the opportunities of transferring these repetitive tasks to the service assistants. In contrast to the classic perception where occupations are encouraged to hive-off “dirty work” for professional status (Hughes 1958; Huising 2015), the regular staff in the study perceive value in doing dirty work. Instrumental incentives for doing dirty work, such as increased autonomy or self-regulation, do not matter as much as in the study by Huising (2015). Rather, in the present study, dirty work seems to be a resource for recovery, which is not worth hiving-off to reinforce a professional image.

That said, there are also interdependencies amongst the factors, and it is empirically difficult to separate factors from one another. They are not mutually exclusive, as they are interwoven in the empirical data. One might observe complementarity; for example, when regular staff are working side by side with the service assistants, this may strengthen their sense of solidarity. Apart from the interdependencies, the strength of the approach is that it reveals six mechanisms that are appropriate to understand how the division of labour is influenced, although without neatly distinguishable causal effects.