Providing individualized services under complex conditions: A configurational analysis of street-level organizations
Abstract
enIndividualized services are provided under complex conditions, as a variety of factors can affect the ability of a street-level organization to adapt its services to individual needs and circumstances. Especially challenging are tensions between the means of control and standardization following new public management (NPM) and post-NPM ideas of holistic and coordinated services. Through a fuzzy-set qualitative comparative analysis of Norwegian sector-spanning street-level organizations, we show three different configurations that can promote individualized services. These consist of variations of structural circumstances (size, service variety); organizational responses (goal coherence, cross working); and manager capacity (professional background, managerial orientation). Service individualization is not an outcome of the interaction between street-level workers and clients alone, but an outcome of street-level organizations and their managers' use of measures and competencies across service sectors, and of their capacity to develop a shared perception of goals and an organization that handles institutional complexity.
Abstrakt
noIndividuelle tjenester leveres under komplekse forhold, ettersom en rekke faktorer kan påvirke førstelinjeorganisasjoners evne til å tilpasse sine tjenester til den enkeltes situasjon og behov. Spesielt utfordrende er spenninger mellom styringsmidler og standardisering knyttet til New Public Management (NPM) og post-NPM ideer om helhetlige og koordinerte tjenester. Gjennom en fuzzy-set kvalitativ sammenlignende analyze av norske førstelinjeorganisasjoner (NAV-kontor) som yter både arbeidsrettede og sosiale tjenester, viser vi tre forskjellige konfigurasjoner som kan fremme individuelle tjenester. Disse varierer når det gjelder strukturelle forhold (størrelse, tjenestespekter), organisatorisk respons (felles målforståelse, arbeid på tvers av sektorgrenser) og lederkapasitet (profesjonell bakgrunn, lederorientering). Tjenesteindividualisering skapes ikke bare gjennom samspillet mellom den enkelte fagperson og klient, men gjennom hvordan førstelinjeorganisasjoner og deres ledere utnytter tiltak og kompetanser på tvers av tjenesteområder, og deres kapasitet til å utvikle felles målforståelse og en organisasjon som håndterer institusjonell kompleksitet.
1 INTRODUCTION
Service individualization (i.e., the process of tailoring services to individual needs and circumstances) is considered a central means by which to increase the responsiveness and effectiveness of public services. To be successful, service individualization requires the autonomy and ability of street-level workers to flexibly use a variety of instruments and apply their professional discretion to adjust services to individual clients (Borghi & Van Berkel, 2007; Rice, 2017; Toerien et al., 2013; Tummers et al., 2015; Van Berkel & Valkenburg, 2007). To the extent that effectively addressing clients' problems requires services from different sectors, organizations and/or professional areas, service individualization requires the integration or removal of barriers between these areas, for example, removal of barriers between employment services and social services to improve the labor market participation of disadvantaged citizens (Heidenreich & Aurich-Beerheide, 2014). Street-level workers' discretion is necessary because their decisions involve judgment and responsiveness to individual circumstances (Brodkin, 2012; Maynard-Moody et al., 2003; Molander et al., 2012), with the consequence that their judgments become, in effect, policy (Lipsky, 2010).
To date, much street-level research has focused on workers' discretion and the interactions between the workers and the citizens (Chang & Brewer, 2022). However, while this individual-level orientation was key to understanding the microlevel provision of service individualization, it tended to neglect how street-level workers operate in circumstances that are not of their own choosing (Hudson, 2014) but rather contingent upon the organizational circumstances in which they work (Rice, 2013). Accordingly, attention has shifted toward the role of street-level organizations (SLOs) and the different governance regimes to which they and their workers are subject (Brodkin, 2011; Brodkin, 2012; van Berkel et al., 2017).
In this article, we extend this research on the organizational conditions for service individualization by drawing on the concept of “institutional complexity,” which highlights how organizations respond to contradictory institutional requirements (Greenwood et al., 2011), particularly those coming from New Public Management (NPM), in its standardization and narrow client orientation, and post-NPM efforts, which focus on cross-sector collaboration and holistic client orientation (Christensen & Lægreid, 2011a; Osborne, 2006; Sørensen & Torfing, 2016). To study combinations of conditions potentially impacting SLO's delivery of individualized services, we apply a configurational methodological approach and fuzzy-set qualitative comparative analysis (fsQCA) (Marx et al., 2013; Ragin, 2008; Thomann, 2019). Accordingly, we formulated the following research question: Which conditions support SLOs in the delivery of individualized services?
In our study, we relied on survey data from managers and workers in sector-spanning SLOs within the Norwegian Labour and Welfare Services (NAV). We focus on structural circumstances (organizational size and service variety), manager capacity (professional background and single-purpose orientation), and organizational means (cross-working and goal-coherence) as conditions at play in these service configurations. By identifying specific configurations of conditions (Rice, 2013; Rice et al., 2018; van Berkel et al., 2017) that can explain service individualization outcomes, our aim is to add to earlier studies of service individualization by following Brodkin's (2012) recommendation to choose an enabling approach and focus on conditions that facilitate quality and responsiveness in policy delivery. Further, we add to the relatively limited research that examines street-level bureaucracy with specific outcome variables related to the client's needs (Chang & Brewer, 2022, p. 2198).
In the following sections, we discuss the perspective of institutional complexity before outlining previous research on service individualization and the conditions under which it is often provided. We then proceed to describe the research setting—that is, the local Norwegian labor and welfare agencies merging two service sectors in partnerships between the state and the municipalities—and continue with a review of our QCA method and an explanation of how we operationalized and calibrated the conditions included in our analysis.
Our findings show three configurations that support individualized services, which can also be understood as organizational responses that successfully combine structural circumstances and managerial capacity to align contradictory institutional demands. Based on our findings, we draw some theoretical conclusions about the conditions supporting individualized services in SLOs.
2 THEORETICAL PERSPECTIVE: PUBLIC ORGANIZATIONS UNDER INSTITUTIONAL COMPLEXITY
A key premise of neo-institutional theory is that organizations and their behaviors are influenced by their institutional contexts. The notion of institutional complexity has been deployed to conceptualize the multiple opposing logics to which organizations must respond (Greenwood et al., 2011; Kraatz & Block, 2008; Skelcher & Smith, 2015). Institutional logics are underlying assumptions or belief systems that shape and define the “rules of the game” and subsequently lend legitimacy to practices in a given setting (Friedland & Alford, 1991; Thornton et al., 2012).
Organizational research has examined the different response strategies adopted by organizations to meet the contradictory demands of different institutional logics (Fossestøl et al., 2015; Greenwood et al., 2011; Kraatz & Block, 2008; Oliver, 1991; Pache & Santos, 2013). At least three overarching responses have emerged in the literature. The first response involves an effort to reduce complexity by adhering to the most dominant logic and ignoring or only symbolically complying with other demands. A second response is to balance demands by compartmentalizing (e.g., creating separate units and finding cooperative solutions between them). Compartmentalization may, however, represent an “uneasy truce” between the logics (Reay & Hinings, 2005), thus requiring additional managerial integration work. A third response is to integrate demands, for example, by developing new ways of work (or even new organizational types) or new (hybrid professional) identities that transgress the institutional logics forced upon them.
When organizations respond to institutional complexity, managers' actions are crucial. Street-level managers are overarchingly in charge of service delivery to the local target population and are thus accountable for their organizations' outputs (Gassner & Gofen, 2018). At the street-level, decisions about appropriate job design, task descriptions, and the division of labor for street-level workers are made (van Berkel et al., 2017). According to Gassner and Gofen (2018), street-level managers translate formal policy decisions to daily work, with specific consideration of the impact on the clients. By translating and adapting policies, managers can influence the ways in which street-level workers use their discretion and lead the workers to move toward clients (Davidovitz et al., 2022).
Street-level managers may take up managerial practices that recognize human judgment as essential for effective public services, and a central challenge for management is to improve the support for discretionary quality in street-level workers' practices (Lipsky, 2010). By facilitating knowledge exchange, coaching, and organizational support, managers can guide professionals toward understanding value conflicts, developing reasonable interventions, and justifying actions (Schott et al., 2016). In contexts where there is a lack of guidance as to what a new policy means in practice, street-level workers may adjust the policy to their already existing practice rather than altering the practice according to the new policy, even when they support the latter (Nørup & Jacobsen, 2023). In more constructive environments, managers can protect valuable services for clients from encroachment on the part of administrative bureaucracy (Bakkeli, 2022b). Managers can also facilitate the reinterpretation and rearranging of the professional's identity to support more interprofessional and integrated ways of working (Breit et al., 2018; Reay et al., 2017). As such, street-level managers are not determined by their institutional contexts, as there is always some leeway. Nevertheless, the institutional complexity in which managers operate creates a contested terrain for their change agency, which often seems restricted by government measures and monitoring (Gatenby et al., 2015).
2.1 Service individualization under complex conditions
The governance models of NPM and post-NPM can be conceived of as two central institutional logics influencing service individualization in a wide range of public services (Christensen & Lægreid, 2022). Governance regimes—as institutional logics—rarely replace each other but are instead typically layered (Capano, 2019; Streeck & Thelen, 2009), meaning that SLOs must handle conflicting demands (Christensen & Lægreid, 2011b; Eriksson & Eriksson, 2023; Fossestøl et al., 2015). Illustrative of such layering is the goal of service individualization. While the idea of personalization has an NPM origin in giving citizens (as customers) increased choice of service providers in a (quasi)market, post-NPM ideas of individualization includes coproduction with citizens and cross-sectoral collaboration (Larsen & Caswell, 2022; Osborne et al., 2016; Toerien et al., 2013).
Extensive research on the impact of NPM has revealed detrimental effects not only on street-level workers' practice, but also on the clients and ultimately the political goals set for the services (Brodkin, 2011; Considine et al., 2020a; Nothdurfter & Hermans, 2018). For example, Soss et al. (2011) have demonstrated how performance management practices disciplined the ways that actors at the street level made decisions. In a similar way, Brodkin (2011) showed how performance measurement initiatives shifted administrative costs to clients, favoring “speed over need,” which effectively delegitimized clients' claims for help. However, there is also evidence that a high caseload is a more important determinant of negative performance than performance management mechanisms (van Berkel & Knies, 2016).
Of particular relevance, here are consequences for the ability to achieve individualized services. While service individualization involves post-NPM features of cross-sector collaboration and coproduction with clients, policies of personalization have been implemented within NPM models of (quasi)markets. Findings from the Australian employment sector showed that the marketization of employment services from the 1990s did not result in the desired flexibility, personalization, and tailoring of services but instead in routinization (Considine et al., 2011). Later reforms aimed at re-introducing flexibility and the tailoring of services seem not to have affected the path toward standardization once set in motion, in part due to sharpened activation policies and increasingly detailed purchaser requirements that intensified providers' risk aversion (Considine et al., 2020b). Fuertes and Lindsay (2016) documented the increased standardization of street-level practice in the UK's Work Program toward a narrow focus on motivating clients by ensuring enhanced job searches and a quick return to work at the expense of individualization.
In the field of healthcare, Eriksson and Andersson (2023) have demonstrated how a service logic ideal was practically unreachable in a context infused by NPM because of the institutionalized focus on the control, measurement, and efficiency of intra-organizational processes. In contrast, Davidovitz et al. (2022) showed that, even in marketized service provision, street-level workers moved “toward clients” rather than away from them. In a related vein, Rice's (2017) study of governance mechanisms aimed at achieving service individualization in the field of active labor market policy showed that in contrast to procedural standardization, institutionalized discretion implied that frontline workers tended to look more broadly for instruments that “worked,” and that this enabled frontline workers to distinguish their service practices for both vulnerable clients and more “regular” work-ready clients. Furthermore, Bakkeli (2022a) showed that even procedural standardization, in the form of standardized employment programs, may be open to individualization if the street-level workers have sufficient discretionary autonomy to prioritize the voice of the client and are actually responsive to the client's wishes.
This above-mentioned research establishes that SLOs may have different ways of managing the complexity involving post-NPM and NPM governance models. Some SLOs have been shown to be primarily oriented toward NPM ideas and act in line with hierarchical steering and control, while others develop strategies of hybridization that may positively combine conflicting demands (Fossestøl et al., 2015).
2.2 Individualization through integrating services
When merging or establishing partnerships across service sectors to create more individualized services, efforts of post-merger integration of organizational cultures, identities, operational tasks, and practices are required (Graebner et al., 2017). Birkinshaw et al. (2000) argue that the success of mergers is contingent on the effective management of “task integration” (i.e., the transfer of capabilities and resources to create value), and “human integration” (i.e., the creation of a shared identity among the workers). One effective means of task integration may be the establishment of cross-functional teams consisting of workers with different professional skills and competences, a high degree of mutual interdependence, and the ability to collaborate to reach the desired objectives (Parker, 2003).
Regarding human integration, managers must integrate different occupational identities connected to frontline workers' expertise and boundaries of jurisdiction (Abbott, 2014; Bucher et al., 2016; Kellogg, 2014). Furthermore, human integration may involve developing new occupational identities (Reay et al., 2017), shared objectives, shared mental models and organizational identity—the shared sense by organizational members of who they are as a group (Graebner et al., 2017, p. 32). Knuth and Larsen (2010) showed the importance of this human dimension, as they found that differences in professional cultures between social workers and employment service caseworkers hampered coordinated client treatment in Danish job centers.
Managers and their professional competence play a key role in post-merger integration. When aiming to change established ways of working, managers have been shown to draw on their embeddedness in the organization, to use their knowledge of a particular setting to make change happen, and to legitimize new ways of working (Reay et al., 2006). Managers may be professionals themselves, serving as “hybrid managers” in boundary positions who combine and coordinate the ideas and concerns of senior management and professional practice (Burgess & Currie, 2013; McGivern et al., 2015; Noordegraaf, 2015). Some professionals in managerial or supervisory positions may take up such hybrid positions willingly, others more reluctantly (Klenk, 2020; McGivern et al., 2015). When combining the processes of involvement and engagement, managers' change leadership can enable reforms to become embedded at the micro level (Higgs et al., 2023). However, street-level managers may differ in their capacities to change the professional identities of frontline workers, capitalize on reform signals, or obtain resources for service innovation (Breit et al., 2018; Breit et al., 2022).
The development of individualized services at the street level is not only a function of the manager's capacity and agency or the available organizational means, but also contingent upon the structural circumstances in which the SLOs operate, as assumed by the idea of congruence (Andrews et al., 2016; Andrews & Boyne, 2014; Donaldson, 2001). One key circumstance is connected to the tasks performed by the organization. According to Rice (2017), individualization depends on sufficient service variety. At the same time, increased service variety also requires increased dependency on coordination mechanisms (Okhuysen & Bechky, 2009). Furthermore, the size of organizations will also affect the need for coordination; while smaller organizations are more likely to involve mutual adjustment and agile, flexible, and collaborative forms of work, larger organizations are generally more open to specialization in one way or another (Mintzberg, 1979). Hence, service individualization will depend on how the complexity of demands and expectations is managed in SLOs.
While all the above-mentioned factors may play key roles individually, there is assumedly not just one single successful response strategy, but rather several feasible responses (van de Ven et al., 2013). Therefore, we may expect that SLOs not only differ in how they handle complexity in implementation of reforms (Fossestøl et al., 2015) but also that responses may involve different configurations of conditions (Mellberg et al., 2023; Zimmerman et al., 2016). Accordingly, a methodological approach focusing on the possible configurations of conditions that may support service individualization is required.
2.3 Study setting
Service individualization is high on the agenda of many governments' efforts to aid in the employment of vulnerable groups (Rice, 2017; Van Berkel & Valkenburg, 2007), and countries across Europe have implemented reforms aimed at promoting individualized support by increasing coordination between the employment and social service sectors (Askim et al., 2011; Minas, 2014).
Our subject of study was the Norwegian Labour and Welfare Administration (NAV). NAV stands out as one of Europe's most ambitious service integration reforms (Champion & Bonoli, 2011; Minas, 2014). At the local level, the reform merged two previously separate service sectors in a partnership between the state's public employment services and the municipalities' social services. In SLOs, the reform introduced an institutional complexity. The state administration brought in an NPM logic of performance measurement, standardized service provision, single-purpose organizations focused on the clients' employment, and an orientation primarily toward the most job-ready clients (Breit et al., 2018; Christensen & Lægreid, 2011b; Jantz et al., 2015). The municipal social services were dominated by a logic of professionalism, that is, a strong social work profession with a professional identity and a tradition of professional autonomy (Breit et al., 2018; Fossestøl et al., 2015; Røysum, 2013). The reform policy was more of a post-NPM whole-of-government logic of tailor-made and comprehensive service provision, with the development of multi-purpose organizations and local autonomy following the self-government of the municipalities (Christensen & Lægreid, 2011).
Beyond the implementation period of gradually merging SLOs (from 2006 to 2011), these multiple contradictory logics have provided a context of institutional complexity for SLOs (Fossestøl et al., 2015; Røysum, 2013). The ambition and comprehensiveness of the Norwegian reform, aimed at target groups with multiple needs, as well as the merging of employment and social services within the organization as more than just “flanking” services (Rice, 2017), makes it a particularly informative research setting for studying the organizational conditions that promote service individualization.
3 RESEARCH DESIGN
3.1 The rationale for using QCA
To study combinations of conditions potentially impacting SLO's delivery of individualized services, we apply a configurational methodological approach and fsQCA (Marx et al., 2013; Ragin, 2008; Thomann, 2019). Fuzzy set QCA is a set-theoretic method with foundations in Boolean algebra and fuzzy set extensions. The method enables the study of social phenomena as sets and to study in what way different social phenomena relate to each other in terms of set relations (Oana et al., 2021, pp. 6–7). Conditions and outcomes are conceptualized as sets to which cases belong. QCA is based on the idea that a condition might be related to the outcome only in combination with other conditions and on equifinality, which means that an outcome may have several explanations, that is, be the result of different combinations of conditions (Oana et al., 2021, p. 8). The notion that sets are fuzzy means that a case can belong to a condition (or set) partially or fully, enabling an analysis that account for the gradualness or degree of a social phenomenon.
3.2 Case selection and scope conditions
The analysis draws on a 2019 survey of managers and workers in 23 SLOs. As the analysis had an explorative purpose, the case selection was related to theoretical propositions and not to populations or universes. With a “diverse case” logic (Seawright & Gerring, 2008), the cases were selected to represent differences in terms of the extent of service individualization, internal integration, manager background, geographic location, and size. Based on the literature presented above, we included the following conditions in our analysis.
3.2.1 Service individualization
Service individualization is operationalized in our analysis as the street-level workers' judgment of whether the organization had achieved its goal of delivering services adapted to individual needs and circumstances. This was our outcome condition.
3.2.2 Organizational size
Throughout the implementation of the reform process, the size of SLOs consistently emerged as a condition impacting internal integration (Fossestøl et al., 2015). Previous research has shown that the larger the organization, the less internal integration across previous service sectors. For example, in 2014, almost 80% of employees in smaller organizations in NAV reported success in finding solutions for individuals with complex needs, compared to only 50% in the largest organizations (Fossestøl et al., 2014). Organizational size is, to some extent, a reflection of the size of the population served by the organization (i.e., small organizations indicate small municipalities). Because many Norwegian municipalities are small, some organizations had fewer than 10 employees, and only the larger municipalities had more than 100 employees at the time of our study. We assume that service individualization is more difficult to achieve in large organizations. Size was assessed by the number of full-time equivalents.
3.2.3 Service variety
Rice (2017, p. 469) has argued that the availability of “flanking social services,” such as counseling, and “housing support and social housing” enhances the potential for service individualization. When an organization provides a social service in-house, it is presumably easier to provide tailor-made assistance for clients. At the same time, if a variety of social services is provided by an SLO, this increases the task complexity to be managed (Andrews & Boyne, 2014). Service variety was assessed by the number of social services included in the SLOs.
3.2.4 Managerial background
Previous research has indicated that “hybrid managers,” with a background as professionals, may use this experience to influence other professionals (McGivern et al., 2015; Reay et al., 2017). Hybrid managers may have greater credibility among professionals than non-hybrid managers, which allows them to promote central features of professionalism while simultaneously creating more integrated forms of work and professional identities (Breit et al., 2018). Accordingly, we assessed managerial background in terms of the managers' educational background in social work.
3.2.5 Managerial single-purpose orientation
Previous research has shown diverging opinions on whether, in line with NPM ideas of single-purpose functionality, the SLOs should standardize their services and limit their target group to the most job-ready unemployed or do the opposite, which would mean including many services to address the multiple problems of groups further away from employment (Breit et al., 2018; Fossestøl et al., 2015). We assumed that managers' opinion that offering many social services obstructs the achievement of the employment goal signals that they align with an NPM orientation who prioritizes “job-ready” clients without multiple needs. We labeled this condition “managerial single-purpose orientation.”
3.2.6 Cross-working
Cross-working is a strategy for task integration (Birkinshaw et al., 2000) in SLOs. This means that caseworkers work across different areas of expertise and across state and municipal service sectors to serve clients' needs. We operationalized this indicator as the degree to which municipal employees reported working in state service areas and vice versa.
3.2.7 Goal-coherence
Human integration (Birkinshaw et al., 2000) requires that cultures and occupational identities from formerly separate services and conflicting demands and expectations be reconciled in a worker community characterized by mutual trust and shared identity (Graebner et al., 2017, p. 32). In our study, an indication of whether human integration was developed was the extent to which the caseworkers experienced coherence between state and municipal goals and believed that the state and municipal goals converged in a shared direction in the SLOs' work.
3.3 Set calibration
Set calibration is the process of assigning membership scores to cases based on raw data, which involves the analytical decision of whether a case belongs to a set or not. This is done to establish the empirical score that best determines the qualitative “difference in kind,” which is referred to as the “crossover pointְ.” We used all three basic types of sets, including crisp, four- and five-value, and fuzzy sets (Oana et al., 2021, pp. 31–33; Ragin, 2008; Schneider & Wagemann, 2012).
Five of the original variables (i.e., raw data) were ordinal and measured with Likert scale scores ranging from 1 to 4 or 1 to 5. Three variables were coded as dichotomous, including the manager's main professional background, the number of social services, and the work-readiness orientation of the managers. We also included an indicator of size based on the number of full-time equivalents in 2017. No joint register of the total number of municipal and state employees existed, and since the size data were from 2017 and those of the survey from 2019, this indicator was an approximate measure of the organization's size.
For the Likert scale variables, we calculated the averages1 as rough indicators of the organizational level. We applied theory-guided calibration (Oana et al., 2021; Schneider & Wagemann, 2012) to recode the variables into fuzzy sets, grouping the responses into qualitatively defined categories of fuzzy-set memberships (cf. Table 1).
| Label | Survey question | Original range | Calibration |
|---|---|---|---|
| “Service individualization” | To what degree services are adapted to individual needs and circumstances |
1. Not at all 2. To a lesser degree 3. To some degree 4. To a great degree |
Fuzzy set: theory-guided 4: fully in (1) <4 and ≥3.5: more in than out (0.67) <3.5 and ≥3.00: more out than in (0.33) <3: fully out (0) |
| “Managerial social work background” | Street-level managers' main educational background: social or healtha | 0–1 |
Crisp set (0–1) 1: fully in (1) 0: fully out (0) |
| “Managerial single-purpose orientation” | Street-level manager's opinion about whether many social services weaken or strengthen the goal of employment |
1. Weakens considerably 2. Weakens some 3. Neither/nor 4. Strengthens some 5. Strengthens considerably |
Crisp set (0–1) 1–2: fully in (1) 3–5: fully out (0) |
| “Service variety” | Number of social services in the street-level organization |
0. Only minimum solution 1. 1–2 services 2. 3–4 services 3. 5 or more services |
Crisp set (0–1) 2–3: fully in (1) 0–1: fully out (0) |
| “Large organization” | Number of full-time equivalents in the street-level organization in 2017 | 9.9–169.5 |
Fuzzy set: direct method 120: fully in (0.95) 70: crossover point (0.5) 40: fully out (0.05) |
| “Cross-working” |
Mean of: 1. Municipal employees work in state job/function areas 2. State employees work in municipal job/function areas |
1. Completely disagree 2. Partially disagree 3. Neither/nor 4. Partially agree 5. Completely agree |
Fuzzy set: theory-guided 5: fully in (1) <5 and ≥4.5: almost but not fully in (0.9) <4.5 and ≥4.00: more in than out (0.67) <4.0 and ≥3.50: more out than in (0.33) (0) <3: fully out |
| “Goal coherence” |
Mean of: 1. Joint perception of goals across state and municipal employees 2. State and municipal goals pull in the same direction |
1. Completely disagree 2. Partially disagree 3. Neither/nor 4. Partially agree 5. Completely agree |
Fuzzy set: theory guided 5: fully in (1) <5 and ≥4.5: almost but not fully in (0.9) <4.5 and ≥4.00: more in than out (0.67) <4.0 and ≥3.50: more out than in (0.33) (0) <3: fully out |
- a Only one manager in our cases had a health education; we therefore speak of the managers as having a social work education/background.
Oana et al. (2021, p. 38) suggest a good area of applying the recoding method to calibrate fuzzy sets is when comparing survey responses on Likert scale variables. The Likert scale variables ensure meaningful variation between the cases that we recoded. The fuzzy sets indicate our interest in focusing on not only on the degree to which some cases perform better than others, but also the distinctiveness between cases of the same kind. This approach captures differences in kind as well as degree.
The differences in coding (4-point fuzzy scores vs. 5-point scores) between the outcome condition at the one hand and “Cross-working” and “Goal coherence at the other, was based on theoretical considerations.” It is reasonable to assume that caseworkers and managers at least sometimes perceive that they succeeded in individualizing services. We therefore set the thresholds for “service individualization” as somewhat more conservative for this condition compared to the other two conditions (corresponding to “to some degree,” on average”). We used 0.9, 0.67, and 0.33 as threshold values as one way of coding among others (e.g., 0.75 and 0.25), capturing differences on both kind and degree. Beyond ranking, the distance does not have a numerical meaning (Oana et al., 2021, p. 38th).
Using the direct method, the threshold for full membership in the condition “Large organizations” was set to 120 full-time equivalents, with the crossover point as 70 and with threshold for full non-membership as 40 full-time equivalents. Our assessment of “large organizations” was based on considerations about the number of full-time equivalents that might require more complex ways of organizing the work. Moreover, earlier Norwegian research (Fossestøl et al., 2020) have categorized NAV offices with full-time equivalents above 70 as “Large.”
The analyses were performed using Rstudio (version. 4.2.3) using the functions from the SetMethods (Oana et al., 2021) and QCA packages (Dusa, 2018). For analyses before determinations of sufficiency, see Tables A1–A4, including descriptive analyses, diagnostics, and analysis of necessity.
3.4 Limitations
The case selection was guided by purpose (as described above) but also by the character of the survey data. Because we were interested in conditions characterizing SLOs that could only be identified by the managers, we selected organizations where the managers had answered the survey. Also, since we aggregated the data by averaging individual workers' responses within each organization and treating them as (rough) characteristics at the organizational level, we needed to focus on organizations with a minimum of eight respondents with valid responses to the variables in our analysis. In the final QCA analysis, we included 23 street-level case organizations that fulfilled these three criteria.
A limitation of the data is lack of information about the response rate in each case organization. However, a low response rate is not necessarily an indicator of poor survey quality (Keeter et al., 2006). Important though, is whether there is reason to believe that the themes addressed systematically influenced the willingness of certain groups of workers to respond. Since the themes addressed were largely uncontroversial, this is not probable. In addition, our analysis, including our outcome measure, was based only on self-reported data. Ideally, our outcome variable should have been service users' perceptions of whether they received individualized services, but unfortunately, no such data existed. Nevertheless, as our analysis shows, there was no uniform tendency among the workers to positively overrate the outcomes of their work.
3.5 Analysis
We performed an analysis of “sufficiency,” that is, whether cases sharing a condition or a combination of conditions also share an outcome (Ragin, 2008). In this analysis, instances of combinations of conditions are considered subsets of instances of the outcome. The analyses of “necessity”—that is, the assessment of whether cases that shared the outcome also shared a condition—are shown in Table A4.
In the analysis of sufficiency, we adopted a truth table, which lists the possible combinations of conditions and outcomes associated with each combination and assesses whether the conditions are consistently related to the outcome (Oana et al., 2021, p. 13). The goal of the truth table analysis is to find the shortest expression of the conditions that are sufficient for the outcome by eliminating irrelevant conditions via logical minimization (Oana et al., 2021, pp. 13, 87). Based on fuzzy set logic, each case can have only at least 0.5 membership in one combination (Ragin, 2008). A case's membership score in a configuration was calculated by taking the minimum membership score for all conditions within the configuration.
The analyses involves several parameters of fit (consistency, proportional reduction of inconsistency [PRI], and coverage) for necessity and sufficiency that allowed us to quantify the extent to which set relations deviated from being “perfect” (Oana et al., 2021, p. 16). Consistency scores indicate the degree to which cases in a specific configuration display the outcome (Ragin, 2008). PRI is an assessment of whether a combination of conditions is sufficient for both the presence and absence of the outcome, and raw coverage is a measure of how much of the outcome is covered (or explained) by a combination of conditions (Ragin, 2008), whereas unique coverage involves assessing the unique or exclusive contribution of a combination.
The minimization process was based on a cut-off value at the consistency level. Combinations with raw consistency scores above the chosen cut-off value were classified as fuzzy subsets of the outcome. We applied a frequency threshold of 1, which is recommended when there are few cases, with a consistency threshold of 0.8, and a PRI threshold of 0.51. The truth table shows that there was a drop in the consistency scores for the included combinations from 0.905 to 0.670 for the combination not included in the minimization process that justified this choice (cf. Table 2), since perfect set consistency is rare. The logical minimization of the truth table also involved identifying and eliminating logically redundant prime implicants (all truth table rows sufficient for the outcome), meaning elements that were not needed to explain the occurrence of the outcome were excluded to prevent the Boolean formula from becoming overly complicated (Oana et al., 2021, pp. 112–113).
| Managerial social work background | Managerial single-purpose orientation | Service variety | Large organization | Goal coherence | Cross-working | OUT | N | Sufficiency inclusion score | PRI: Proportional reduction in consistency | Cases |
|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 1 | 0 | 0 | 1 | 1 | 1 | 2 | 1 | 1 | 13, 16 |
| 1 | 0 | 1 | 1 | 1 | 0 | 1 | 1 | 1 | 1 | 1 |
| 1 | 1 | 0 | 0 | 1 | 1 | 1 | 3 | 0.909 | 0.816 | 17, 19, 23 |
| 0 | 0 | 0 | 0 | 1 | 1 | 1 | 3 | 0.905 | 0.861 | 3, 10, 11 |
| 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 0.670 | 0 | 20 |
| 0 | 0 | 0 | 0 | 0 | 1 | 0 | 3 | 0.662 | 0.333 | 7, 9, 18 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.637 | 0 | 21 |
| 1 | 1 | 1 | 1 | 1 | 0 | 0 | 1 | 0.534 | 0 | 6 |
| 0 | 1 | 0 | 0 | 0 | 1 | 0 | 1 | 0.493 | 0 | 22 |
| 1 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0.493 | 0 | 4 |
| 1 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.493 | 0 | 5 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.451 | 0.130 | 8 |
| 0 | 0 | 1 | 1 | 0 | 0 | 0 | 2 | 0.419 | 0.009 | 2, 15 |
| 0 | 1 | 1 | 1 | 0 | 0 | 0 | 2 | 0.274 | 0 | 12, 14 |
Depending on which remaining rows were included in the logical minimization, we obtained different solution formulas. We blocked out the possibility of including untenable assumptions (i.e., counterfactuals making statements on the conditions for the opposite of the outcome) and conducted an “enhanced standard analysis” as proposed by the literature (Oana et al., 2021, p. 130; Schneider & Wagemann, 2012).
We show the results for the enhanced conservative solution, as we were mostly interested in exploring the complexity of solutions in cases in which no remainders were taken into account. This was based on the assumption that combinations without empirical observations are not sufficient for the outcome (Oana et al., 2021, p. 124). The results for the enhanced parsimonious solutions and enhanced intermediate solution are shown in Tables A5 and A6. The Appendix also includes robustness checks following the protocol published by Oana et al. (2021) to ensure the reliability of the QCA results (cf. Tables A7–A11). Based on these analyses our results seem robust.
To better understand and explain the various configurations identified, we applied survey data to further describe cases that illustrated the configurations.
4 RESULTS
The analyses of necessity showed that none of the conditions (i.e., either their presence or absence) passed the consistency threshold of being necessary for individualization (cf. Table A4). The analysis of sufficiency included six conditions that could be either present or absent, resulting in 64 (2K = 26) logically possible combinations. Only 14 rows had empirical cases, illustrating the problem of limited diversity (cf. Table 2).
There was one “deviant case consistency in kind” that had membership in the combinations that exceeded its membership in the outcome of “individualization” (X > 0.5 and Y < 0.5), which contradicted our sufficiency claim (Case 11). In addition, the results showed that we were unable to explain the presence of individualization for Case 18, which would be referred to as a “deviant case coverage” in the literature (X < 0.5 and Y > 0.5) (Oana et al., 2021, pp. 92–97).
The results for the conservative solution showed three configurations. The configuration covered by most cases (five) showed that a combination of service managers not finding service variety (within the organization) to weaken the aim of employment, in combination with service variety, not being a large organization, having a joint goal and direction between state and municipal workers, and working across the employment and social services, explained individualization (cf. Table 3). This configuration characterized SLOs with a high variety of social services, and accordingly, the managers in these cases did not find service variety to weaken the overall aim of employment, which relates to workers' experiences of goal coherence and thus having a shared understanding of the organization's goals and mission comprising their occupational identity.
| Configurations | inclS | PRI | covS | covU | Cases | |
|---|---|---|---|---|---|---|
| 1. ~managerial single-purpose orientation * service variety * ~large org. * goalcoh. * crossw. | 0.945 | 0.881 | 0.385 | 0.385 | 13, 16, 17, 19, 23 | |
| 2. ~managerial social work background * managerial s.p.o * ~service variety * ~large org. * goalch. * crossw. | 0.905 | 0.861 | 0.184 | 0.184 | 3, 10, 11 | |
| 3. managerial social work background * ~managerial s.p.o * ~service variety * large org. * goalch.* ~ crossw. | 1.000 | 1.000 | 0.066 | 0.066 | 1 | |
| M1 | 0.938 | 0.884 | 0.635 |
- Note: M1: ~managerial single-purpose orientation * service variety * ~large organization * goal-coherence * cross-working + ~managerial social work background * managerial single-purpose orientation * ~service variety * ~large organization * goal coherence * cross-working + managerial social work background * ~managerial single-purpose orientation * ~service variety * large organization * goal coherence * ~cross-working > service individualization.
- a The notation ~ indicates absence of the condition, and the asterisk (*) indicates AND, meaning the minimum value of the combination.
Case 16 is an organization illustrating this configuration, located in a municipality at the south-west coast of Norway with almost 14,000 inhabitants. The municipality is situated nearby one of the largest cities in Norway which employ almost one third of the inhabitants in the municipality. Most of the respondents from this municipality evaluated the collaboration with flanking services as fairly good. The manager was employed by the state, had former management experience from the state administration and was experienced, having served as a leader for five or more years.
The configuration covered by the second most cases (three) showed that the combination of a manager without a social work education with a single-purpose orientation, in conjunction with the absence of service variety, not being a large organization, having a joint goal and direction between state and municipal workers, and working across the employment and social services, also explained individualization (cf. Table 3).
Case 10 illustrates the second configuration. This organization was in a rural municipality with almost 30,000 inhabitants nearby a large city. The manager was employed by the state, had former management experience in state administration, and had served as a manager for a few years. Most of the workers evaluated the collaboration with flanking services as good.
Both the first and the second configuration included the conditions of having a joint goal and direction between state and municipal workers and working across employment and social services. The most parsimonious solution indicated that the combination of goal coherence and cross-working were the core conditions (cf. Appendix, Table A5). Important differences between the two configurations are that the second configuration entails a street-level manager mostly oriented toward NPM ideas promoted by the state administration as well as the absence of service variety.
Both configurations in the first and second configuration were characterized by the absence of a large organization (i.e., below 70 full-time equivalents). The cases had between 10 and 42 employees, except for one case in the second configuration, Case 11, with 68.6 full-time equivalents.
Case 11 was a “deviant case consistency in kind,” being more out than in the set of service individualization, which meant the case contradicted the claim of sufficiency for the second configuration, perhaps due to being a fairly large organization. Sensitivity analyses with lower anchor points for size removed case 11 from the configuration (results not shown). However, as the analyses showed more complex configurations for both the intermediate and parsimonious solutions, we decided to keep the existing anchors. While Case 11 had a manager employed by the state with experience in the state administration, the manager had served as a manager for only 1 year, which meant that the former manager's orientation (which was not known to us), probably more than the current one, had influenced the street-level workers' perceptions of the services to the clients.
The configuration covered by the least number of cases (one) had a manager with a social work education in a large organisation who did not find service variety to weaken the goal of employment. There was also a joint goal and direction between the state and municipal employees. However, there was an absence of service variety and working across the employment and social services (cf. Table 3).
Case 1, an organization located in a large municipality with about 60,000 inhabitants, not far from the capital of Norway, illustrates this configuration. Most of the respondents evaluated collaboration with flanking services as good. The manager was employed by the municipality and had experience from both social services and the private sector. The manager was experienced and had served as a manager for five or more years.
Taken together, the solution coverage (0.635) showed that our solution formula did not account for all fuzzy-membership scores in the outcome. Case 18 was more in than out of the set of “service individualization” but was not explained by any solution. This indicates that there were other combinations of conditions of importance that might have explained individualization not included in our study. Case 18 was a small organization with fewer than 20 employees located in the southern part of Norway in a municipality with just above 6000 inhabitants. Most of the respondents evaluated collaboration with flanking services as good, which is a possible explanation for the positive outcome score.
5 DISCUSSION AND CONCLUSION
In this article, we have examined the configurations of conditions through which SLOs provide individualized services to target groups with multiple needs (Rice, 2017; Tummers et al., 2015). Our point of departure was that individualized services are provided in complex institutional environments, meaning that managers need to balance and navigate different institutional demands (Greenwood et al., 2011; Kraatz & Block, 2008). A variety of conditions potentially affect SLOs' ability to adapt their services to individual needs and circumstances, of which tensions between NPM standardization and control and post-NPM ideas of holistic and coordinated services are among the most difficult to converge in the integration of sector-spanning services.
With our QCA comparison of 23 SLOs, we join Hupe and Buffat (2014) in supporting an “agenda for street-level bureaucracy research with a more systematic comparative logic” (p. 549) that captures variation in the structural circumstances and organizational conditions for street-level workers' practice. Our analysis shows that in some SLOs workers believed that their organization was able to individualize services adapted to the clients' needs—meaning that the organizations had been able to manage contradictory institutional demands in meaningful ways—but also that in other SLOs workers believed that they did not. Of the 23 studied SLOs, only in seven of these, workers strongly believed that the organization delivered services adapted to individual needs and circumstances. Our analysis confirms previous research showing that organizational conditions vary, and that how they are managed matters (Breit et al., 2018; Fossestøl et al., 2015; Mellberg et al., 2023; Zimmerman et al., 2016). Thus, there are several configurations of managerial capacities and organizational responses that may generate service individualization. The analysis revealed three configurations that can promote individualized services. The first suggests that in small organizations, the provision of individualized services can be facilitated by internal service variety. Such a provision is underpinned by high internal integration, which is itself produced through post-merger task integration (cross-working) and a belief among workers about shared goals and direction (human integration; conf. Birkinshaw et al., 2000). Apparently, these smaller organizations have managed to integrate the contradictory institutional demands involving NPM and post-NPM to provide individualized services, as opposed to adhering to one demand or compartmentalizing them within the organization (Fossestøl et al., 2015; Greenwood et al., 2011; Kraatz & Block, 2008; Oliver, 1991; Pache & Santos, 2013; Reay & Hinings, 2009).
This ability to integrate demands suggests the presence of a manager who (a) has facilitated a reinterpretation and rearranging of the street-level workers' occupational identity and stimulated more interprofessional and integrated ways of working, and (b) is positive about including a high service variety that might better serve citizens with multiple needs. It further suggests a manager who has utilized leeway for local discretionary judgments and protected valuable services to clients by buffering against the detrimental effects of NPM, which entail standardization and a single-purpose orientation toward only employment-oriented services. In these organizations, the manager's education does not seem decisive; rather, the manager's non-acceptance of a single-purpose organization is what makes a difference here. This indicates that although a social work education might spur an orientation toward serving clients with multiple problems further away from the labor market, managers with other educational backgrounds may also develop such an orientation. Presumably, individualization is stimulated in such SLOs by the presence of flexible work forms, where frontline workers can profitably and flexibly supplement each other's competences and capacities. Likewise, smaller SLOs enable mutual adjustment (Mintzberg, 1979), which is a crucial coordinating mechanism whereby workers communicate informally with each other. In smaller organizations, such adjustments are likely less inhibited by physical structures, such as different floors or buildings.
The second configuration suggests that individualized services can also be provided in smaller organizations through high internal integration, but with a manager seeing high service variety as an obstacle to the organization's employment goal. This seemingly contradictory configuration, involving no service variety, has at least two possible explanations. The first is that street-level workers and managers in such contexts believe that the organization should provide individualized services to a more limited group of the most work-ready citizens. Hence, they operate with a narrow perception of individualization. This explanation suggests that SLOs have managed the conflicting institutional demands by eliminating demands from one of the service sectors.
The other explanation is that individualized services are provided not in-house but through extensive collaboration with flanking services (as in Case 10). In this way, the narrow, single-purpose conception in the organization is complemented by other services. This way of providing individualized services relies on compartmentalizing responses by drawing boundaries between the SLO and other service providers and finding cooperative solutions between these organizations. At the same time (as Case 10 illustrates), management shifts may reduce the manager's influence on the street-level workers, meaning that their outcome evaluations are shaped by their own practices rather than the manager's opinions.
The third configuration shows that service individualization can also be accomplished in large organizations, which are more prone to specialization and less susceptible to flexibility and mutual adjustment (Mintzberg, 1979). This specialized way of working is evident in a context with low service variety and low cross-working among groups of frontline workers. At the same time, internal specialization is presumably integrated by a hybrid manager with a professional background and thus professional expertise, credibility, and social capital (Klenk, 2020; McGivern et al., 2015; Reay et al., 2017). Assumedly, this background enables the manager to perceive crucial commonalities across specializations and explain them in meaningful ways to frontline workers. This feature is evident in the street-level workers' perception of the organization as having shared goals and direction.
This third configuration signifies an SLO that meets conflicting expectations by compartmentalizing demands and creating separate units responding to each of them. A hybrid manager's professional perspective and legitimacy among the street-level workers may have prevented compartmentalization from producing partial rather than holistic and individualized services, although with limited internal service variety. The description of case 1 showed that the street-level workers seemed able to collaborate effectively with other service providers, and since this was a large municipality, we can assume a richer variation in flanking social services than in smaller municipalities, which may further contribute to service individualization. Finally, this configuration thus shows another meaningful way of managing institutional complexity, which is by compartmentalizing conflicting demands through the creation of separate departments—in our case separating between state and municipal services. In this way, larger SLOs can compensate for their lack of flexibility by relying more on the advantages of specialization.
Our study emphasizes that individualization cannot be understood merely as an outcome of the interaction between a street-level workers and clients alone, which has received most of the attention in street-level research (Chang & Brewer, 2022). Rather, individualization is promoted by organizing work processes so that street-level workers engage in tasks across the involved service sectors (cross-working) and have a shared perception of goals that transcend former occupational identities and cultures. Hence, individualized services can be construed not only as the result of willing and competent frontline workers or targeted social policies or reform efforts, but also as an outcome of how SLOs and their managers make use of measures and competencies across different service sectors and institutional demands.
Our study thus emphasizes that management matters (Lipsky, 2010; Reay et al., 2017; Schott et al., 2016), especially street-level management, which encompasses the upper echelon of the frontline organization. Such organizations have the organizational means at hand to design and develop work processes and influence workers' occupational identities, but also constraints entailed in contradictory demands from NPM and post-NPM ideas.
Overall, these findings add to street-level studies that, according to Brodkin (2012, p. 945), “virtually up-end the classic management control versus street-level autonomy dichotomy” by turning the attention to the complex organizational conditions that enable street-level work toward individualized services (Rice et al., 2018). Our study is novel in emphasizing the influence of a post-NPM logic, which has been considerably less researched in the street-level literature than NPM steering. Post-NPM ideas of cross-sector collaboration and coordinated and individualized services in street-level work entails shifting the circumstances for street-level workers and managers (Chang & Brewer, 2022). Our study suggests that service individualization may be accomplished in SLOs influenced, among other factors, both by NPM and post-NPM, and that this requires a heightened capacity on the part of the street-level manager to (strategically) develop the organization to handle this complexity.
We believe that the findings regarding different configurations to service individualization that depend on the orientation and capacities of street-level managers, the size of the organization, the internal specialization and coordination, and the number of services it is required to deliver, are theoretically valid. However, these findings need to be empirically investigated across different empirical contexts. Our findings underscore the recent emphasis on how the NPM-inspired standardization of service tasks hinders service individualization (Fuertes & Lindsay, 2016; Rice et al., 2018), and the role of street-level managers in mediating policy and influencing frontline workers' practices (Gassner & Gofen, 2018) and in responding differently even in similar national governance contexts (Fossestøl et al., 2015; Mellberg et al., 2023; Zimmerman et al., 2016). Likewise, we assume that goal coherence across service sectors is no less important elsewhere. Thus, we have some reason to suggest that similar dynamics are likely to be rather generic traits of frontline development and the provision of individualized services.
At the same time, our analysis was conducted in a Norwegian context, with a universal, social democratic welfare regime which to a lesser extent than many other Western countries, has been dominated by NPM steering, and with a mostly enabling activation policy. This context differs from the marketization, standardization, and narrow activation focus characterizing Anglo-Saxon countries such as the United Kingdom and Australia (Considine et al., 2020a; Fuertes & Lindsay, 2016). Furthermore, in Norway, rather than relying on cooperation between different service agencies the employment and social services have been merged into one SLO through a partnership model, which has been considered among the most radical reforms in Europe (Champion & Bonoli, 2011). This has rendered much complexity to be managed within SLOs since withdrawal from cooperation is not possible, which may have placed stronger pressure on both managers and workers than if they were involved in more voluntary cooperation with other agencies. Hence, the factors studied here might not have the same strength in more voluntary contexts.
Furthermore, because services that otherwise would have been “flanking” services delivered by other SLOs were incorporated into some of the organizations studied here, we might assume that requirements for manager capacity to handle such service variety are somewhat higher here than in organizations with lower service variety. At the same time, the higher service variety included in smaller SLOs seems to offer good conditions for service individualization, a finding that may be relevant in contexts other than the Norwegian one. Research from the United Kingdom, Germany, and Italy has demonstrated that underdeveloped organizational interfaces between employment and social service providers tend to constrain individualization to the disfavor of vulnerable jobseekers (Rice et al., 2018).
To conclude, while service individualization in general takes place under complex conditions, these conditions, as well as how they are managed, vary among SLOs. When such complex conditions are appropriately managed, service individualization rather than standardization is produced, albeit through different configurations. Acknowledging that there are different configurations means that standardization is not a way toward individualizing services, neither at the level of service provision nor at the level of policy reforms.
ACKNOWLEDGMENTS
All authors have contributed equally to this article. The analysis is based on survey data collected by the Work Research Institute at OsloMet. The authors would like to express their gratitude for the opportunity to apply these data.
FUNDING INFORMATION
This work was supported by the Research Council of Norway, Grant Numbers 301943 and 269298.
CONFLICT OF INTEREST STATEMENT
The authors declare that they have no conflict of interest.
Endnote
APPENDIX
| Service individualization | Managerial social work background | Managerial single-purpose orientation | Full time equivalents | Service variety | Cross-working | Goal coherence |
|---|---|---|---|---|---|---|
| Min.: 2.55 | Min.: 0.00 | Min.: 0.00 | Min.: 9.90 | 0.00 | Min.: 2.14 | Min.: 2.21 |
| 1st Qu.:3.26 | 1st Qu.: 0.00 | 1st Qu.:0.00 | 1st Qu.: 23.80 | 0.00 | 1st Qu.:3.42 | 1st Qu.:3.53 |
| Median: 3.39 | Median: 0.00 | Median: 0.00 | Median: 41.20 | 0.00 | Median: 4.14 | Median: 3.81 |
| Mean: 3.39 | Mean: 0.30 | Mean: 0.35 | Mean: 57.37 | 0.44 | Mean: 3.93 | Mean: 3.82 |
| 3rd Qu.:3.63 | 3rd Qu.:1.00 | 3rd Qu.:1.00 | 3rd Qu.: 76.80 | 1.00 | 3rd Qu: 4.55 | 3rd Qu.:4.26 |
| Max.: 4.00 | Max.: 1.00 | Max.: 1.00 | Max.: 169.50 | 1.00 | Max.: 4.96 | Max.: 4.66 |
| Cases | fsService individualization | Managerial social work background | Managerial single-purpose orientation | Service variety | fsLarge office | fsGoal coherence | fsCross-working |
|---|---|---|---|---|---|---|---|
| 1 | 0.67 | 1 | 0 | 0 | 0.75 | 0.67 | 0 |
| 2 | 0.33 | 0 | 1 | 0 | 1 | 0.33 | 0 |
| 3 | 1 | 0 | 1 | 0 | 0.01 | 0.90 | 0.9 |
| 4 | 0.33 | 1 | 0 | 0 | 0.01 | 0.33 | 0.67 |
| 5 | 0.33 | 1 | 1 | 0 | 0.15 | 0.33 | 0 |
| 6 | 0.33 | 1 | 0 | 1 | 0.71 | 0.67 | 0.33 |
| 7 | 0.33 | 0 | 0 | 0 | 0.01 | 0.33 | 0.67 |
| 8 | 0 | 0 | 1 | 0 | 0.16 | 0 | 0.33 |
| 9 | 0.33 | 0 | 0 | 0 | 0.01 | 0.33 | 0.67 |
| 10 | 0.67 | 0 | 1 | 0 | 0.06 | 0.67 | 0.67 |
| 11 | 0.33 | 0 | 1 | 0 | 0.47 | 0.90 | 0.67 |
| 12 | 0.33 | 0 | 0 | 1 | 0.84 | 0 | 0 |
| 13 | 0.67 | 0 | 0 | 1 | 0.02 | 0.67 | 0.90 |
| 14 | 0 | 0 | 0 | 1 | 1 | 0 | 0 |
| 15 | 0.33 | 0 | 1 | 0 | 0.98 | 0 | 0 |
| 16 | 0.67 | 0 | 0 | 1 | 0.02 | 0.9 | 0.67 |
| 17 | 0.67 | 1 | 0 | 1 | 0 | 0.67 | 0.90 |
| 18 | 0.67 | 0 | 0 | 0 | 0.01 | 0.33 | 0.90 |
| 19 | 0.67 | 1 | 0 | 1 | 0.05 | 0.90 | 0.90 |
| 20 | 0.33 | 0 | 0 | 1 | 0.20 | 0.33 | 0.33 |
| 21 | 0.33 | 0 | 0 | 0 | 0.05 | 0 | 0 |
| 22 | 0.33 | 0 | 1 | 1 | 0.02 | 0 | 0.67 |
| 23 | 0.67 | 1 | 0 | 1 | 0.01 | 0.67 | 0.9 |
|
Calibrated 4-value fuzzy set, theoretical calibration—“Service individualization” Skewness Cases >0.5/Total number of cases: 9/23 = 39.13% Calibrated crisp set—“Managerial social work background” Skewness Cases >0.5/Total number of cases: 7/23 = 30.43% Calibrated crisp set—“Managerial single-purpose orientation” Skewness Cases >0.5/total number of cases: 8/23 = 34.78% Calibrated crisp set—“Service variety” Skewness Cases >0.5/total number of cases: 10/23 = 43.48% Calibrated fuzzy set, direct method—“Large organization” Skewness Cases >0.5/total number of cases: 6/23 = 26.09% Calibrated 5-value fuzzy set, theoretical calibration—“Cross-working” Skewness Cases >0.5/total number of cases: 13/23 = 56.52% Calibrated 5-value fuzzy set, theoretical calibration—“Goal coherence.” Skewness Cases Cases >0.5/total number of cases: 10/23 = 43.48% |
| Cons.Nec | Cov.Nec | RoN | |
|---|---|---|---|
| Managerial social work background | 0.356 | 0.524 | 0.828 |
| Managerial single-purpose orientation | 0.322 | 0.415 | 0.762 |
| Service variety | 0.453 | 0.467 | 0.709 |
| Large organization | 0.286 | 0.451 | 0.821 |
| Goal coherence | 0.829 | 0.862 | 0.905 |
| Cross-working | 0.766 | 0.713 | 0.789 |
| ~ Managerial social work background | 0.644 | 0.416 | 0.428 |
| ~ Managerial single-purpose orientation | 0.678 | 0.467 | 0.500 |
| ~ Service variety | 0.547 | 0.435 | 0.576 |
| ~ Large organization | 0.876 | 0.549 | 0.468 |
| ~ Goal coherence | 0.615 | 0.486 | 0.596 |
| ~Cross-working | 0.571 | 0.494 | 0.648 |
- a The notation ~ indicates absence of the condition.
| Configurations | inclS | PRI | covS | covU | Cases | |
|---|---|---|---|---|---|---|
| 1. ~service variety*goal coherence | 0.889 | 0.735 | 0.441 | 0.129 | 3,10,11, 1 | |
| 2. goal coherence *cross-working | 0.927 | 0.838 | 0.701 | 0.389 | 13,16, 3,10,11, 17,19,23 | |
| M1 | 0.915 | 0.804 | 0.829 |
- Note: M1: ~service variety*goal coherence + goal coherence *cross-working− > service individualization.
In the enhanced parsimonious solution, any remainders that resulted in a logically simpler solution (regardless of theoretical and substantive knowledge), were taken into account (Oana et al., 2021, p. 126). The enhanced intermediate solution included easy counterfactuals, in line with our theory based directional expectations about single conditions (Oana et al., 2021, p. 123). We expected that the presence of “managerial social work background,” absence of “large organization” and absence of “managerial single-purpose orientation” in combination with presence of “service variety,” “goal coherence,” and “cross-working” would explain the outcome.
| Configurations | incIS | PRI | CovS | covU | Cases | |
|---|---|---|---|---|---|---|
| 1. ~large organization*goal coherence*cross-working | 0.944 | 0.872 | 0.697 | 0.665 | 13,16,3,10,11, 17, 19, 23 | |
| 2. managerial social work backg.* ~ managerial single-purpose or.* ~ service variety* goal coherence | 1.000 | 1.000 | 0.097 | 0.065 | 1 | |
| M1 | 0.948 | 0.884 | 0.762 |
- Note: M1: ~large organization*goal coherence*cross-working + managerial social work background* ~ managerial single-purpose orientation* ~ service variety*goal coherence−> service individualization.
Robustness test protocol
The enclosed robustness checks follow the protocol published by Oana et al. (2021) to ensure the reliability of the fsQCA results.
Sensitivity ranges
|
Exclusion: Lower bound NA Threshold 40 Upper bound 65 Crossover: Lower bound 45 Threshold 70 Upper bound 85 Inclusion: Lower bound NA Threshold 120 Upper bound NA |
- a Adding and subtracting five employees in a maximum of 30 rounds (the maximum runs specified).
For the 0.5 crossover threshold, the Boolean formula for the sufficient solution remains the same no matter where we place the threshold within the lower bound of 45 and the upper bound of 85. For the inclusion threshold, adding a 5-step value 30 times, the software could not find an upper bound where the solution changes (Oana et al., 2021, p. 153).
| Raw consistency T.: | Lower bound 0.67 | Threshold 0.8 | Upper bound 0.9 |
| N. Cut: | Lower bound 1 | Threshold 1 | Upper bound 1 |
Table A8 shows that the Boolean formula of the solution stays the same only when the raw consistency threshold is placed within the 0.67–0.9 range and when the frequency cut-off is fixed at one case (Oana et al., 2021, p. 153).
Fit-oriented robustness
| RF_cov | RF_cons | RF_SC_minTS | RF_SC_maxTS | |
|---|---|---|---|---|
| Robustness_Fit | 0.994 | 0.991 | 0.776 | 0.753 |
Table A9 shows the comparison of various parameters of fit of the IS the robust core (RC) and the minimal and maximal test. All parameters range from 0 to 1 where higher values indicate robustness. The results show that the solution is quite robust in terms of fit measures when tested against a series of plausible analytic changes. The values of RFcov and RFcons are high, indicating substantial overlap. The values of RFSC_minTS and RFSC_maxTS are lower, but indicate a reasonable degree of set coincidence between the initial solution (IS), the minTS, and the maxTS (Oana et al., 2021, p. 155).
| RCR_typ | RCR_dev | RCC_Rank | |
|---|---|---|---|
| Robustness_Case_Ratio | 1 | 0.5 | 2 |
Table A10 shows that of all typical cases, all cases are members of both IS and minTS and, therefore robust. Of the deviant consistency cases, 50 per cent are robust for both IS and minTS. The last part of the output (Rob_Case_Rank) takes the value of 2 indicating that we simultaneously have cases that fall into both categories “shaky” and “possible” (Oana et al., 2021, p. 157). Table A11 shows the names and cases for each type; robust typical, robust deviant, extreme deviant, and irrelevant cases.
|
Robust Typical Cases (IS*MIN_TS and Y > 0.5): Cases in the intersection/total number of cases: 8/23 = 34.78% Cases in the intersection/total number of cases Y > 0.5: 8/9 = 88.89% Case Names: 1 3 10 13 16 17 19 23 Robust Deviant Cases (IS*MIN_TS and Y < 0.5): Cases in the intersection/total number of cases: 1/23 = 4.35% Cases in the intersection/total number of cases Y < 0.5: 1/14 = 7.14% Case Name: 11 No Shaky Typical Cases (IS* ~ MIN_TS and Y > 0.5): No Shaky Deviant Cases (IS* ~ MIN_TS and Y < 0.5): No Possible Typical Cases (~IS*MAX_TS and Y > 0.5): Possible Deviant Cases (~IS*MAX_TS and Y < 0.5): Cases in the intersection/total number of cases: 1/23 = 4.35% Cases in the intersection/total number of cases Y < 0.5: 1/14 = 7.14% Case Names: 6 Extreme Deviant Coverage Cases (~IS* ~ MAX_TS and Y > 0.5): Cases in the intersection/total number of cases: 1/23 = 4.35% Cases in the intersection/total number of cases Y > 0.5: 1/9 = 11.11% Case Names: 18 Irrelevant Cases (~IS* ~ MAX_TS and Y < 0.5): Cases in the intersection/total number of cases: 12/23 = 52.17% Cases in the intersection/total number of cases Y < 0.5: 12/14 = 85.71% Case Names: 2 4 5 7 8 9 12 14 15 20 21 22 |
Open Research
DATA AVAILABILITY STATEMENT
Research data are not shared.
