Frontline Professionals Performing Collaborative Work with Low-Income Families: Challenges across Organizational Boundaries

This article discusses certain challenges relating to interagency collaboration between the Norwegian Labour and Welfare Administration (NAV) and Child Welfare Services (CWS). We have asked what obstacles to holistic work with lowincome families who receive measures from NAV and CWS simultaneously can be identified. The departure point is collaboration on a local project at the municipal level. The differences between the views of the individual services (and the mandates based on these views) with regard to parental obligations have proved challenging. Using the theory of institutional logic, we have explored how different logics have influenced these services’ approaches to parenthood and the significance of these influences for interagency collaboration. We have also investigated how caseworkers1 in the two services have managed to create 1 In Norway, NAV and CWS use different terms for the frontline workers in the services. We have chosen to use “caseworker” in this article.


Introduction
Although Norway is a rich country, an increasing number of children are living in relative poverty (UNICEF, 2016). To reverse this trend, policy-makers and researchers have called for holistic approaches that can be used by the various welfare services responsible for these children and their families (Fløtten & Grødem, 2014;Langeland, Dokken, & Barstad, 2016;Malmberg-Heimonen, Tøge, Rugkåsa, Fossestøl, Liodden, Bergheim, Gyüre & Buzungu, 2019; Ministry of Children, Equality and Social Inclusion, 2015; The Governmental Board and Health Supervision, 2013; The Office of the Auditor General in Norway, 2013Norway, -2014. In this article, we focus on challenges related to a project called "New ways of working in interagency collaboration with low-income families," aimed at developing and improving collaboration between the Norwegian Labour and Welfare Administration (NAV) and Child Welfare Services (CWS) in this area on a local level.
When parents and children in low-income families receive assistance from NAV and CWS simultaneously, this indicates that the families are economically disadvantaged and that there are concerns about the children's care situation.
These children and their parents have diverse needs related to their individual family circumstances, and these needs require multiple measures across organizational boundaries. However, welfare services for children and families with poor finances and challenging life situations are rarely integrated or coordinated; they often focus on separate problems or individual members of the family. Therefore, new forms of collaboration and guidelines to improve the assistance provided to this group of children and their parents are required (Ask & Sagatun, 2015;Gustavsen, Meij, Nilsen & Braathen, 2012;Malmberg-Heimonen et al., 2019;NAV & Bufdir, 2016;Oterholm, 2018).
NJSR -Nordic Journal of Social Research Vol. 11, 2020 We identify some of the specific challenges relating to collaboration across these two services by highlighting how different approaches to parenthood have influenced efforts to develop coordinated support. We took a trailing research approach, whereby we followed the project over two years (INNOS 2013(INNOS -2015 Ask & Sagatun, 2019). The data for our analysis mainly consist of statements from and discussions among frontline caseworkers and middle managers from NAV and CWS. We regard the study as a contribution to raising awareness of important issues that can be difficult to recognize and name in the day-to-day practical work of these services.

Theoretical Framework
Welfare services handle many problems that can be characterized as "wicked problems" (Rittel & Webber, 1973). These are complex problems that do not have simple unambiguous solutions (Vabø, 2014, p. 17). They span different areas of expertise. Thus, effective solutions may require collaboration across services, which is the case with our subject of study. The theory of institutional logics can help us to understand certain challenges that services face in addressing such problems. Alford and Friedland (1985) first introduced institutional logics as an approach for exploring the interrelationships and contradictory practices and beliefs inherent in the institutions of modern Western societies. Since then, the theory has been developed further, and greater emphasis has been placed on the interactions between actors, organizations, and institutions (Thornton & Ocasio, 2008;Thornton, Ocasio, & Lounsbury, 2012). The perspective entails an awareness of the socially constructed, historical patterns of practices, assumptions, values, beliefs, and rules that give meaning to a certain social reality. It concerns a set of presumptions and perceptions embedded in a particular field that guide the actions of the social actors in this field (Thornton & Ocasio, 2008;Thornton et al., 2012). Freidson (2001) has described organization/bureaucracy, market, and profession as ideal types of organizing businesses and has examined how 2 The research was part of the main project Innovation and Service Development through Evolving Forms of Collaboration (INNOS, 2013(INNOS, -2017, partly funded from The Research Council of Norway. these constitute contexts with different institutional logics that guide and control practice and practitioners. Social classification and categorization are key mechanisms by which institutional logics might shape individual cognition (DiMaggio, 1997). For example, the supports to be given to certain citizens or users are assessed in the context of the categories the respective institution, in this case NAV or CWS, perceives as its responsibility.
The concept of an institution is ambiguous; it covers both general categories, such as the family and the state, and concrete businesses and organizations.
The term is also used more generally to define enduring structures that create stability and meaning and shape and regulate behaviour (Scott, 2008).
Regulative, normative, and cultural-cognitive elements are the central building blocks of institutional structures (Scott, 2008, p. 49). CWS and NAV are organizations with overall institutional arrangements; they have defined purposes related to legislation and current policy, and they contain guidance based on their social mandates and social obligations for actions and interactions (Oterholm, 2018;Scott, 2008;Thornton et al., 2012). Institutions' regulative aspects, which may include sanctioning activities, can be either informal or formalized and assigned to specialist actors (Scott, 2008). The regulative aspects of NAV and CWS are examples of the latter.
According to Thornton et al. (2012, pp. 76-77), social actors are key to understanding institutional persistence and change. Dominant institutional logics should not be understood as specific scripts for action but rather as core principles for organizing activities and channelling interests. This understanding relates to the concept of embedded agency (Thornton et al., 2012). Battilana and D'Aunno (2009) argue that although structures and institutional pressures tend to reproduce the status quo, forms of institutional work and agency can also challenge institutions. The concept of institutional work provides a broader view of agency by highlighting the intentional and practical actions through which institutions are created, maintained, and disrupted, often through relatively mundane and ordinary activities (Lawrence, Suddaby, & Leca, 2009, p. 1).
We use institutional logics to call attention to the fact that NAV and CWS address parenthood on different grounds due to their institutional rationales as welfare organizations in separate policy areas. However, they have overlapping responsibilities on certain issues, such as the multifaceted problems faced by low-income families. These problems run across various sectors of society. By studying how caseworkers in these different services talk about their practices, we highlight both important barriers and the institutional work that has been performed to overcome those barriers to collaboration. The latter corresponds to interagency cooperation, which, inter alia, is about finding common ground and recognizing mutual dependency across institutional boundaries when dealing with complex social problems (Gray, 1989;Willumsen, 2009). Our analysis is inspired by these theoretical perspectives. Based on our approach, we developed the following research questions: How can we understand the categorizations of parenthood in NAV and CWS? How have these categorizations influenced practice in relation to low-income families using both services? How might categorizations and subsequent practices affect obstacles to and opportunities for achieving a holistic approach to collaboration between NAV and CWS?
Existing Research on the Institutional Logics of NAV and CWS NAV is itself the result of an interorganizational reform that brought different agencies with historically different orientations together: the rule-orientated social security system, needs-orientated social services, and the resultsorientated employment office (Andreassen & Fossestøl, 2014, p. 177).
Referring to Freidson (2001), Andreassen and Fossestøl (2014) argue that the prevailing logic within NAV is dominated by organization, bureaucracy, and hierarchy. Oterholm (2015) interviewed social workers in CWS and NAV to study whether and how the institutional framework for their professional practices affected differences in judgements regarding youths in need of support when leaving care. She found that social workers in NAV were mainly influenced by a public logic in relation to a "generalised other," while social workers in CWS, although influenced by a public logic, were led by a private, or family-orientated, logic in relation to a "concrete other." During her fieldwork in a CWS office, Vagli (2009Vagli ( , 2014 found that social workers predominantly used emotional language in their internal conversations and judgements relating to the categorization of children and parents for available measures. She had expected a more bureaucratic logic to appear and questions whether CWS is sufficiently aware of its institutional embeddedness in a particular cultural and historical context. Without discounting emotions or elevating bureaucracy, she points out that it is a challenge for CWS, on both institutional and individual levels, to be aware of and balance different logics. Within a UK context, Morris, White, Doherty, and Warwick (2017, p. 53) have argued that child welfare takes place in a moral context where parents are constructed as potentially culpable for problems exhibited by the child. This is particularly apposite when applied to judgements about parenting. They claim there has been a gradual shift in how families and parents are perceived and named in child welfare services. They are no longer seen to be struggling in the face of adversity but rather presented as wilfully failing to exercise good judgement, seize opportunities, and work hard. They are expected to collaborate to change their own situations, and less attention is paid to their social and economic circumstances. An increasing focus on risk assessments influences how families are seen and spoken about (Morris et al., 2017).
The common wisdom that can be taken from these studies is that organization and institutional affiliation matters, especially with respect to socially constructed systems of categorization that influence practice (Thornton et al., 2012). In relation to our study, NAV seems to be characterized by a bureaucratic and public logic, while the logic of CWS seems more blurred and more influenced by a private logic where specific parental obligations are at stake.
We recognize that different institutional logics can exist inside welfare services and can compete for dominance. Agents in the field experience and express ambivalence in their professional practice.
We emphasize that our study focuses on how different internal categorizations, in this case, connected to perceptions of parenthood, affect collaboration between two welfare services and how this influence is reflected in practice.

Materials and Method
The project in which we have been involved was initiated by a local NAV office, and CWS was invited to participate. The project was part of an established partnership between NAV and the university, and it developed out of common research interests and development work (Sagatun, 2013). The research was reported to the Norwegian Centre for Research Data (NSD).
Our methodological approach was trailing research, which defined our positions as researchers who contributed to clarifying potential room for action and supporting development and new ways of doing things (Patton, 2011). We also involved the opinions of partners from practice, thereby creating arenas for dialogue. We remained open-minded regarding their contributions at all stages of the research process (Uggerhøj, 2012).
The involvement of families receiving benefits from both services was crucial to the project, and 10 families gave their consent to participate (Ask & Sagatun, 2019) 3 . Due to the scope of this article, direct citations from our interviews with the families are not prominent in the text. The main source material for this article consisted of interviews with caseworkers and further discussions and feedback from NAV and CWS employees during jointly organized events for the project.
Twenty caseworkers were interviewed in 10 pairs. Each pair consisted of one caseworker from each service, and both caseworkers in each pair dealt with the same family. The caseworkers were interviewed separately from the families with the consent of the parents involved.
All interviews were taped and transcribed, although the transcriptions were not verbatim. Summarized accounts of the interviews with the families and caseworkers formed a basis for further discussion in two workshops and a seminar. The participants of these workshops were the caseworkers and certain middle managers from both services. The first workshop was an open session, and the second workshop was organized so that groups (of mixed NAV and CWS personnel) moved between stations and were given predetermined themes to discuss. The discussions were taped and transcribed. A seminar was also conducted, involving external and internal initiators, as well as a mixed group of employees from NAV and CWS, totalling about 40 people.
The purpose of these events was to allow a space for participants to share, explore, and discuss their experiences and to identify and facilitate new forms of professional practice. We acted as interpreters and facilitators of conversations between participants with differing institutional contexts and, consequently, different starting points and sometimes different languages (Uggerhøj, 2012).
We recognize that researchers are actors themselves. Many years ago, we, the authors, worked in the child protection area. Subsequently, we worked on development and research projects, especially in NAV. Therefore, we are both insiders and outsiders (Hammersley & Atkinson, 1995). This gave us advantages and disadvantages regarding access to data, potential to influence responses, and bias in data interpretation. However, we see our methodological approach and our participation in the larger research group in the project Innovation and Service Development through Evolving Forms of Collaboration (INNOS) as strengthening the validity of the research (Ask & Sagatun, 2019).

Data Analysis
To analyse the data derived from the interviews and workshops and the written reports from the seminar, we used thematic analysis (Thagaard, 2013). By paying attention to how the employees from NAV and CWS described their tasks and the need for collaboration, we identified how they talked about parenthood and expected parenting practices. This emerged as an important theme in the analysis. We examined their stated reasons and explanations for the prevailing approaches in relation to their institutional context and compared the statements of actors in the two services. We focused on the main patterns and nuances in these statements.

Limitations
One challenge we faced was a high turnover in personnel. A lack of continuity and general time pressures in the services may have weakened the engagement of the actors and limited the effectiveness of the trailing research approach. Another limitation of our data is that we did not study the services' evaluations or decisions recorded in the parents' or children's records.

Practice Relating to the Conceptions of Parenthood and Dominant Institutional Logics of NAV and CWS
The policy of "workfare," with "the work line" (arbeidslinja) and slogans like "work first," characterize the official rhetoric of NAV. Generally, this is a statement of consensus rather than controversy (Ministry of Labour and Social Affairs 2012, 2015). As a point of departure, these policies provide a direction for individual agency in NAV (Andreassen & Fossestøl, 2014;Røysum, 2013), including in caseworkers' meetings with adults who, for various reasons, are out of work, dependent on benefits, and parents of minors.
CWS ensure that children who are living in conditions that represent a risk to their health or development receive the help they need when they need it. This includes providing support to parents (NOU, 2016: 16). "The child's best interest" is the organization's guiding and legitimating principle. However, this principle represents normative issues. Research has highlighted that CWS's mandate of intervening in and regulating family life is based on values, power, knowledge, and special ways of interpreting children's needs, parents' obligations, and the service's mission (Morris et al., 2017;Oterholm, 2018;Vagli, 2009).
To explore whether and how various aspects of institutional logics were communicated and how they may have influenced actors, we present examples of how the caseworkers described and reflected on their practices, especially in terms of approaches to parenthood.

Practice in NAV in Relation to Low-Income Parents
Several statements highlighted the fact that the caseworkers were not expected to speak directly with the children. The primary task of NAV is to map the parents' situations and to categorize them to ensure the correct interventions.
This approach subordinates the position of children.
For some, their caseload contributed to this approach: "We who work with AAP [work assessment allowance] have 220 users each." Another caseworker reported that "AAP is very adult-orientated; we are not supposed to consider the children." Yet another caseworker elaborated on this: At NAV, where the work line is the guiding principle, I think it is invasive to talk about children. I do not find it natural to talk about kids, to ask the parents about the children's situation. In the survey form for employability assessment, there is no space to write anything about the situation of children.
The following quotation substantiates this point: "We only focus on-or mostly focus on-measures such as work and activity." In a broader sense, children were considered to be affected by the parents' situations, and the work line was regarded as benefiting them: "To get parents into work can be good for the children." However, such observations were rarely linked to specific reflections on how the child was influenced by the parent's or parents' unemployment and the family's economic situation.
From these comments from NAV employees, it is reasonable to conclude that families and children are viewed from within a narrow frame in the prevailing logic at NAV. Especially when caseworkers must manage an extensive number of cases, they are forced to adjust to the external demands of the work line and map a rather limited picture of the families that focuses on the grown-ups.
In general, the data indicate that workfare, as an embedded guideline, limits the caseworkers' approaches to parenting and causes them to prioritize initiatives that support work and other activities necessary to achieve self-sufficiency. We see that categorization of service users into predefined categories within the system influenced further mapping and led to the implementation of standardized measures. The term embedded agency (Thornton et al., 2012) may be too strong as some caseworkers explicitly claimed that such institutional conditions limited their ability to provide adequate support to families and restricted their individual agency and their exercise of professional judgement (Freidson, 2001). However, these conditions represented an obstacle to holistic and flexible cooperation across NAV and CWS.

Practice in CWS In Relation to Low-Income Parents
"The child's best interest," which is the overall guideline in CWS, seems to be connected to "care first" as the expected direction for caseworkers to take in their approach to parenthood. In our data, this assumption is expressed in several statements from caseworkers, such as the following: "In our work, the focus is only on care and assessing whether it is good enough." Caseworkers use legislation, professional knowledge, internal discussions, and various other methods to assess whether a parent, in relation to a child or children, represents a "good" parent, a "bad" parent, or a parent who "could do better," to use the words of Morris et al. (2017, p. 53).
During the first workshop, one caseworker said, Why am I here? I do not have time to deal with these issues, although I recognize they are important. My obligation is to secure the best interests and welfare of the child, so I must prioritize other measures [measures other than cooperation with NAV].
In one example reported by a caseworker, a mother with severe financial problems asked for assistance from CWS, partly because she had experienced a conflict with NAV. However, CWS's attention immediately turned to what the service defined as the mother's mental health problems and how they affected the children: "So we set aside the economic issues." According to the caseworker, the mother felt cheated and was despairing.
A statement from a middle manager at CWS expresses a similar sentiment: "I think that financial support is not relevant to the care of children. You can be a good caregiver despite a low income. Care is about attachment and relationships, not about how much money you have." Caseworkers from CWS also expressed objections to including the family's caseworker from NAV in the family's supervisory group (ansvarsgruppe). The argument made by CWS was that such inclusion would likely give too much attention to economic conditions at the expense of other important factors concerning the children's care situation.
The data with the examples and quotations above indicate that CWS does not perceive that concrete economic and material conditions lie within their area of responsibility and regards these conditions as peripheral to parenting. We can view these perceptions in the light of CWS's overall institutional conditions and regulative and normative elements (Morris et al., 2017;Scott, 2008;Vagli, 2014). Altogether, such approaches decrease the scope of action across organizational boundaries and highlights limitations and blind spots in CWS's perceptions of parenthood.

Institutional Logics
Most of the caseworkers and middle managers in NAV and CWS involved in the project were social workers by profession. Based on this, we could have expected that they would have a common professional logic. We found traces of this, but the main pattern in our data indicates that the impact of their different institutional contexts is profound (Freidson, 2001).

NAV-Parents as Breadwinners
The data, including the comments presented above, suggest that the overall institutional logic of NAV is dominated by an organizational and bureaucratic structure. As such, it fits the picture that similar studies have drawn (Andreassen & Fossestøl, 2014;Oterholm, 2015). In our study, this logic translates to a view of parents as mainly breadwinners.
Parents become breadwinners through paid work, which is the prioritized and desirable way of overcoming child poverty (Langeland et al., 2016). The mission of NAV is to get people into work or other activities that will enable them to achieve economic self-sufficiency. This goal is mainly operationalized through standardized and rule-based working methods (Røysum, 2013). These standardized methods are meant to ensure that NAV users receive equal and fair treatment. Aiming to provide equal treatment in this way can, however, be difficult to reconcile with a desire to provide each family with assistance tailored to their specific requirements. This is a pertinent issue because it may take many parents in the target group a long time to achieve the goal of paid work.
The parents in these families often experience health problems and social problems over time. The view of employment as beneficial to people's health supports the view of work as the overarching solution (Ministry of Labour and Social Affairs, 2015). Ultimately, there are expectations that parents will adapt to the requirements, rules, and regulations embedded in the institution, according to which users who receive benefits should be motivated to work.

CWS-Parents as Caregivers
Compared to the bureaucratic logic of NAV, we found that CWS was dominated more by a professional logic in the sense that the caseworkers often referred to theoretical perspectives where they largely emphasized psychological knowledge concerning children's development and needs. Similarly, they focused on whether parents were aware of and able to attend to their children's needs.
However, their approach seems to lead to an understanding of parents as mainly caregivers and entails a distance to other aspects of parenthood and to families' multifaceted everyday lives. We therefore understand this approach as reflective of a limited professional logic.
Our reason for this interpretation is that the concrete financial situation of the parents did not appear to be mapped and evaluated as especially relevant to the child's care situation. The issue of how financial issues might affect a parent's self-image as a parent and his or her capacity and capability to exercise parental care was overlooked. This conclusion corresponds with findings from Vol. 11, 2020 comparable research (Andenaes, 2004;Kojan & Fauske, 2011;Morris et al., 2017;Vagli, 2009).
The understanding CWS has of its mandate raises the question of how this service should cope with poverty and inequality as a part of children's care situations. Some statements from caseworkers in CWS indicate that paying attention to poverty issues amounts to patronizing parents rather than recognizing the struggles such issues cause in families' daily lives and that it could distract attention from CWS's mission.
Compared to research describing CWS as influenced by a private or familyoriented logic (Oterholm, 2015) and CWS caseworkers' assessments as dominated by an emotional language (Vagli, 2014), our data provide a slightly different perspective. However, we must emphasize that we have investigated issues other than the aforementioned research, and different logics can exist without being mutually exclusive, or they might compete for dominance (Thornton et al., 2012). Thornton et al. (2012) indicate that individuals who are embedded in a particular institutional logic are more likely to invoke knowledge that is part of that logic.

Challenges and Bridging Differences
In many ways, the empirical data we present in the previous sections supports this reasoning. The employees seemed aligned to the categorization of parents as mainly breadwinners in NAV and caregivers in CWS.
On the one hand, it is not surprising that a review of the data shows that caseworkers mainly understood their tasks in relation to families in terms of the official descriptions of their services' goals. As employees, they must be committed to striving to meet these goals. Furthermore, the internal organization of tasks entails guidelines regarding the content and execution of their professional work (Freidson, 2001;Scott, 2008). On the other hand, emphasizing these standards is not incompatible with being open to supplementary and competing approaches, which may disturb dominant institutionalized categorizations (Thornton et al., 2012).
As explained in the section on our methods, the participating employees met during joint interviews, two workshops, and a dialogue seminar. Their statements during these events provide examples of variations and nuances in their thinking that highlight efforts to bridge differences and implement change.
One caseworker from CWS commented, "Work is not only about wages… Children, they feel ashamed when mum or dad does not go to work." This reflection highlights the possible normative impact that the work-line approach has even on children. We also interpret it as a widened perspective on parenting as it brings into play issues that are connected to societal relationships outside the interactions between children and parents. In response to a discussion on data derived from interviews with parents who expressed the view that CWS did not provide sufficient attention to the effects of economic hardship on their parenting practices, a middle manager from CWS asked, "Are we really that narrow-minded?" Some subsequent reflections from caseworkers supported the parents' viewpoint.
Since some CWS caseworkers regarded economic affairs as relevant to their work with families, we can understand these statements as expressing competing perspectives within the same field. Such internal tensions may provide a basis for thinking about how things could be done differently and how individuals can act as "change agents" in the organization (Thornton & Ocasio, 2008), in this case to widen a limited professional logic. Such actions could also prepare common ground for interagency collaboration (Willumsen, 2009).
Our data show that while NAV caseworkers generally agreed with the work line as the guiding principle, they also expressed frustration in this regard. They communicated the view that efficiency requirements and the measurement of how many users were transferred into work or work-like activities every month limited them from following up on parents who had multifaceted problems. This was also seen as an obstacle to obtaining insights into the everyday lives of the children. The following quotation substantiates this claim: "NAV does not pay close enough attention to parents who are struggling. Things slip in a hectic schedule. We do not have the little wiggle room necessary to obtain tailored solutions." This observation supports the findings of other research that shows that social workers are under pressure to work in a standardized and "simplified" way (Røysum, 2013).
In several of the joint interviews with family caseworkers from NAV and CWS, the NAV caseworkers claimed that the conversation helped to expand their view of the parents' situation. In one example, as a result, a mother was moved to a category that granted her an improved level of follow-up from her caseworker in NAV. Prior to the conversation, NAV had placed her in a predefined category with standardized efforts that required minimal follow-up regarding a work placement. Similarly, one caseworker from CWS said, "Cooperating with NAV is so important for illuminating other aspects of a parent's situation." One NAV caseworker commented, "If we had more knowledge, more information about what they are struggling with . . . Maybe we push too hard?" Because of this broadening of their perspectives, some NAV caseworkers argued that they should themselves follow up with parents rather than outsourcing the task to external actors, which was a widely implemented practice.
Many of the responses during the joint interviews began with statements such as "If we had only known this [earlier], we could have . . ." We interpreted such statements as indicating that a different and better choice or measure could have been implemented. One reflection from a caseworker at CWS underlines this point: "Through working together, we obtain greater knowledge about one another and we can be assured that we gain a partner and a collaborator, not an adversary." A similar view is reflected in another quotation: "We do not have to wait for a problem to talk together." Suggestions were made about sharing expenditures in current cases and in response to parents' requests. A more radical view was to have a common budget for working with these families. The idea of having regular "family meetings" gained significant support, as did making a holistic "family plan" with consent and cooperation from the family in question. The purpose should be to clarify how to provide parents and children with the services they need and are entitled to. Caseworkers commented that the current situation in which comprehensive efforts are lacking results in humiliating experiences for parents who are sent back and forth between the services.
Most of the specific proposals represented small-scale initiatives. While these represent important institutional work (Lawrence et al., 2009), they are not sufficient to change the bigger picture. However, the discussions made the complexity of the issues more apparent, and the caseworkers emphasized that top managers should contribute to establishing more interorganizational collaboration (Willumsen, 2009).
We understand these joint discussions and processes as recognition that different forms of knowledge from multiple actors are necessary to introduce new practices and that changes must be institutionalized to be implemented in practice. From our theoretical approach, we see these views as steps towards disturbing and altering the dominant institutional logics.

Summary and Conclusion
In this article, we deliberately emphasize certain perspectives to clarify what we see as important issues in the discussion on developing a holistic approach to assisting low-income families. Theoretical perspectives can counterbalance simplified answers to complex problems and help to highlight more fundamental challenges that need to be addressed both to understand and to change established practices.
The prevailing logics of both services are embedded in the broader formation of current society (Freidson, 2001;Thornton et al., 2012). Child poverty is a longstanding issue in the public sphere, and although there is cross-political agreement on the need to solve the problem, agreement on effective measures across public sectors is difficult to realize (Fløtten & Grødem, 2014;Malmberg-Heimonen et al., 2019;Vabø, 2014). and CWS can reinforce a narrow perspective that focuses on the individual rather than seeing family poverty from a wider systemic perspective. In our view, this represents a major obstacle to achieving a holistic approach involving collaboration between the services. Measures to address the issues are inadequate and, at worst, can keep families in a poverty trap. We found signs indicating that the services and the caseworkers themselves were trapped by institutional barriers to the extent that they did not see opportunities to expand the scope of their judgements and actions.
The caseworkers had limited room to manoeuvre. Demanding tasks and time pressures limited their ability to challenge the established institutional logic of their organization when necessary. Nevertheless, initiatives to adjust and improve interagency collaboration for specific families were realized and, in general, a more flexible approach emerged over the course of the project. We regarded the interviews and workshops and the seminar as reflexive spaces for negotiation and as essential for an understanding of interdependence (Uggerhøj, 2012;Willumsen, 2009). Barriers built into the structures themselves were acknowledged, although they were not thoroughly problematized. A stronger commitment from leaders was desired.
The project demonstrates that neither the work-line approach nor intensive follow-ups to strengthen parent-child interactions are enough to provide the support low-income families need. The project collaboration was basically an initiative to overcome the shortcomings. Although we see the approaches as potentially complementary in collaborations between NAV and CWS, we argue that both mainly emphasize solutions at the individual level, particularly in relation to parents' attitudes and actions. An extended view is necessary. From our study, we point to reducing bureaucratization and standardization and strengthening a broader professional logic (Freidson, 2001) in and across the services.
Our trailing research was carried out from 2013 to 2015. The statistics show that the number of children living in families with persistently low incomes in Norway further increased after this period (Omholt (ed.), 2019). The challenges are still there. The question of how to meet these challenges has mainly been addressed by local initiatives (Fløtten & Grødem, 2014), including the project present here.
One area for further research could be to explore what a reinforced professional logic might mean for developing comprehensive holistic services for and together with affected families. Another suggestion is to look further into collaborative relationships to unpack the institutional work and frontline agency that might contribute to disrupting organizational boundaries and, eventually, institutional logics.
In the wake of this project, a larger regional development project has been established, which includes more services and cross-sectorial ownership. A research project (2019-2023) with both qualitative and quantitative work packages is generated connected to the new project.
While solutions and measures are needed on multiple levels, welfare services such as NAV and CWS must reach parents and children in low-income families on a local level. Our ambitions with this paper are to contribute to the professional conversation on these complex issues and improve interagency collaboration.