Struggling to counter school segregation- a typology of local initiatives in Sweden

ABSTRACT Swedish compulsory schools are committed to work for equality and social cohesion. Increasing school segregation, however, challenges this commitment. Based on survey data from Swedish municipalities, this article maps and analyses local initiatives that counteract school segregation. We identify three main types of initiatives—reinforcement, dispersal, and merging—and the exogenous (school external) and endogenous (school internal) drivers involved in each of them. The analysis reveals several gaps between the national level, the municipal level and local schools that hamper local efforts to counter school segregation. This article contributes to increased knowledge on how local initiatives of counteracting segregation are constrained by national policies about school choice and independent versus municipal schools, but also how local initiatives tend to focus on organizational dynamics rather than on social and pedagogical processes.


Introduction
Since 1962 when a curriculum for the Swedish compulsory school system was first put in place, one of the main commitments of the Swedish school system has been to promote equality and social cohesion (Ministry of Education, 1962Education, , 1969Education, , 1985Education, , 2010)).This means that compulsory schools are mandated to ensure equal educational and developmental opportunities for children and to counteract the effects of socioeconomic differences within the student body (Pareliussen et al., 2019;SOU, 2020, p. 28).Increasing school segregation has, however, challenged this commitment.Children with similar parental education and socioeconomic and ethnic backgrounds are increasingly clustered within certain schools, and this is reflected in differences in educational opportunities and school results (Andersson et al., 2010;Bunar, 2010;Östh et al., 2013;Voyer, 2019).The differences between children from different schools are widening (Pareliussen et al., 2019), which in turn increases the stratification of Swedish society (SOU, 2020, p. 28).
Increased school segregation in compulsory schools can partially be attributed to societal developments such as residential segregation, rising income inequality, and increased immigration (Andersson et al., 2012;Hansen & Gustafsson, 2016;Hennerdal et al., 2020) but also to conditions created by educational reforms dating back to the early 1990s (Trumberg & Urban, 2021;Yang-Hansen & Gustavsson, 2016).Since the 1990s, the educational system has been characterized by a large independent, for-profit school sector and a strong emphasis on students' and parents' right to choose which school the student will attend (Hennerdal et al., 2020).Meaning that the Swedish educational system has changed from one of the most centralized systems in the world to one of the most decentralized, even if obvious recentralization tendencies have been evident in recent years (Bergh & Wahlström, 2018;Blomqvist, 2004;Grek, 2020;Hennerdal et al., 2020;Musset, 2012;Yang-Hansen & Gustavsson, 2016).These developments have laid the groundwork for an increased stratification of the educational system.Children with a Swedish background, of high socioeconomic status and with well-educated parents more often attend schools of their choice, in other words, these families have the ability and knowledge to navigate the education system to ensure the most favorable outcomes for their children.On the other hand, children who are not privileged tend to have fewer opportunities to navigate the school market for the best schools (Branden & Bygren, 2018;Forsberg, 2018;Hultqvist & Lidegran, 2021).
Until now, scholars have mainly mapped and analysed the causes and consequences of segregation in schools.Less focus has been directed to the local initiatives that seek to counter its negative effects.To address this lacuna, we sent a survey to compulsory school superintendents in Sweden's 290 municipalities.With this data, our aim was to map and analyse local desegregation initiatives.Three research questions are in focus: 1) if and how school segregation is considered a problem in Swedish municipalities, 2) what the municipalities are doing to counter the negative effects of school segregation and 3) what challenges the superintendents encounter in this work.
As a result, the article presents a typology of three main desegregation initiatives, reinforcement, dispersal, and merging, the exogenous (school-external) and endogenous (school-internal) drivers involved in each initiative and the constraints the superintendents' experience in their work to counter the problem.
2. Setting the stage: the transformation of the Swedish educational system Historically, Nordic education systems, including Sweden's, have been built on what is known as the Nordic model of education (Imsen et al., 2017;Lundahl, 2016), which has emphasized the responsibility of the welfare state, universal, equal education, and the role of schools for promoting equity, social justice, democracy, and community building (Antikainen, 2006;Oftedal Telhauga et al., 2006).Since the 1980s, however, this model has been gradually transformed as Nordic countries have turned to a neoliberal management of education, although to different extents (Antikainen, 2006).Education has incorporated a market-driven logic characterized by competition and efficiency, involving the introduction of school choice and the expansion of independent school alternatives (Antikainen, 2006;Imsen et al., 2017).

Decentralization and recentralization: the neoliberal turn and accountability
The marketization of education is not a trend limited to the Nordic countries; it is part of a global restructuring of education, signaling a paradigmatic shift in the governance and organization of education (Grek, 2020;Trumberg, 2019).The many global policy similarities notwithstanding, it is important to recognize differences between countries and how educational policies are shaped nationally and locally (Sivesind et al., 2012).
In Sweden, which is the Nordic country where marketization of compulsory education has been most fully embraced (Musset, 2012;Oftedal Telhauga et al., 2006), this shift has meant that the long tradition of equity and "school for all" stands side by side with increased competition and individual choice (Imsen et al., 2017;Wiborg, 2013).The educational reforms of the 1990s grew out of a criticism of the public sector for being ineffective, costly and bureaucratic, in tandem with a call for education to be more focused on employability.These reforms resulted in a large independent, forprofit school sector, individual choice, and increased competition between independent and public schools (Blomqvist, 2004;Oftedal Telhauga et al., 2006;Wiborg, 2013).This has made Sweden's educational system unique; it is the only OECD country that has a substantial and growing forprofit school sector that is publicly funded (Pareliussen et al., 2019).
The reforms of the early 1990s also shifted a large part of the responsibility for education from the national level to the municipals, including mandatorship for the schools, organizational development, and professional development (Pareliussen et al., 2019;Wiborg, 2013).In 1993 the government deallocated funding that had been earmarked for education, which left the municipalities free to redesignate these resources to municipal affairs, schools or otherwise, as they saw fit (Pareliussen et al., 2019).
Since the 1990s, municipalities have funded both independent and municipal schools.They only, however, have jurisdiction over the municipal schools while companies and/or foundations manage the independent schools.Through a voucher system, where schools are funded per student, students can apply to any school they want, independent or municipal, even outside their home municipality (Hennerdal et al., 2020).Some municipalities give additional funding to schools based on the socioeconomic demographics of the students.However, because municipalities use different funding formulas, the funding of schools is not uniform across the country (National Board of Education, 2011).Admission also varies between public and independent schools.Admission to public schools is based on proximity (the relative proximity rule), where students who live within the school's catchment area have priority (Ministry of Education and Research, 2010).On the other hand, the independent schools often have a waiting list, and if the school is popular, students have to be placed on this list at a young age to ensure their later admission (Trumberg & Urban, 2021;Varjo et al., 2018).
The Swedish system is highly decentralized, but since the early 2000s, there has been a clear tendency to recentralise certain aspects of the educational system; the state has taken a clearer role, through its public institutions, as a control agent within the decentralized system (Hennerdal et al., 2020).This has resulted in a stronger emphasis on measurable results, accompanied by standard-based curriculum reforms, juridical regulations, and increased control and accountability (Bergh & Arneback, 2016;Bergh & Wahlström, 2018).Over the past decades, there have been obvious changes in the balance between local autonomy and national control, with paradoxical expectations often landing at the local level to be handled ultimately by the teachers (Bergh, 2015).Altogether, this means that local authorities and schools are responsible for ensuring national standards of achievement, quality of education, and equivalence but that achievement is measured nationally.

School segregation: a national problem
In this context, it can be concluded that neither local authorities nor schools exist in a vacuum.Especially in recent years, it has become apparent that increased residential delineation and stratification of Swedish society, together with the organizational structure of the Swedish school system and the option to choose schools, have all contributed to school segregation and increasingly uneven academic results (Granvik Saminathen et al., 2018;Hennerdal et al., 2020;Holmlund, 2016;Holmlund et al., 2019;Pareliussen et al., 2019;Söderström & Uusitalo, 2010).These contextual factors highlight how municipalities are limited in their options to define the local school market, while at the same time they are obliged to provide equal education for all their students and enhance the social cohesion of the Swedish education system (Ministry of Education andResearch, 1985, p. 2010).Although the Ministry of Education supports school actors in various ways, the main responsibility for counteracting local school segregation lies with the municipalities.
In municipalities, the pattern of segregation is affected by several factors.For instance, local school markets are determined by the number of schools in the municipality, the relationship between municipal and independent schools, the students' perceptions and choice of schools, and the ability of local school market competition to attract students (Varjo et al., 2018).Given the contingency of several of these factors, planning and organization of local schools are often hard to anticipate.Local authorities are also limited in their ability to manage student school selection (Varjo et al., 2018).In addition to these challenges, it is also important to recognize that the 290 Swedish municipalities have different organizational and funding structures and varying levels of resources available.Important questions for this article, and ones we map, are what strategies and tools municipalities use to counter segregation, the constraints they encounter in this work, and what they consider necessary measures to counter the negative consequences of school segregation.

Data and method
Our interest in analysing different types of local initiatives to counter school segregation emerged from a previous research project where the authors studied one type of initiative-dispersion-in one of Sweden's larger municipalities, Örebro (Arneback et al., 2021).When comparing initiatives in other Swedish municipalities, for example, Nyköping (Kornhall & Bender, 2019) or Malmö (Wigerfelt, 2010), it became clear that there were different views on how to understand and address the problem of school segregation locally (also se Lund, 2021).Therefore, we decided to make the project national and sent out a survey to all municipalities in Sweden with the intent to develop a typology of local initiatives that counter compulsory school segregation and to visualize the possibilities and challenges inherent in each of these processes.
A five-question survey was sent to all compulsory school superintendents in Sweden's 290 municipalities.Given that superintendents work in political organizations and that we wanted the answers to be as honest as possible, the survey was anonymous.Consequently, it was not possible to identify which municipalities the superintendents belonged to.Also, we were not able to use the data to identify certain features of municipalities (size, location, demographics, etc.), although the answers often provided insight into the general characteristics of the municipality.It was also important to have a short, general survey because we expected that long, detailed surveys, given the superintendents' workload, would result in a lower response rate.This type of short survey obviously meant more limited qualitative data, but it allowed us to map general tendencies, understandings of the problem, and local processes of change.
The survey was sent out in January with a return date of March 2021; the response rate was 68% (197 municipalities out of 290).The answers were coded through a thematic analysis, in which we utilized a constant comparative logic (Tie et al., 2019).Inspired by Braun and Clarke (2006), the analysis was done in four steps, with the intention to progressively sort the material into aggregated themes that reflected the main differences in the respondents' answers: 1) We familiarized ourselves with the data by reading and re-reading the answers with the aim to get ideas for initial themes to begin the coding.2) We used the initial codes to capture main themes in relation to the survey questions (see Table 1).Answers that could not be categorized were given special consideration, as they Open answer might form new themes or sub-themes.Any new themes and sub-themes were gradually integrated into the coding scheme through an abductive process.When we were finished with coding and sorting the data, we conceptualized the final themes and sub-themes.3) We noted how common the themes were in the answers.Given that the response rate for each question varied, we grouped the codes into the percentage categories of uncommon (0-20%), common (20-50%) and very common (+50%).In this study, none of the initiatives was very common.We also selected coded quotations that best illustrated each theme and sub-theme.
In some of the quotations we made minor language edits to enhance readability without changing their meaning.4) In order to make visible and analyse the processes inherent in different local initiatives to counter school segregation, we introduced two analytical concepts, exogenous and endogenous drivers.These concepts have their origin in urban theories from the 1990s (Slater, 2013) and have been used previously in research on segregation, in particular, regarding neighborhood effects (Friedrichs et al., 2003;Van Ham et al., 2012).In these studies, exogenous effects are factors that have an impact on the residents but originate outside of neighborhoods, for example, the level of services provided in the neighborhood.Endogenous effects, on the other hand, originate within neighborhoods, for example, the level of social control (Friedrichs et al., 2003).
Drawing on past research, we use exogenous and endogenous drivers to describe efforts that are intentionally used to make other things progress or develop in a certain direction, in this case efforts to counter school segregation and promote equality and social cohesion.The division between exogenous and endogenous drivers made it possible to differentiate between efforts that originated and relied on actors outside of the schools (exogenous) and efforts that used resources already in place in the schools (endogenous).Examples of exogenous drivers are changes in local policies, the involvement of social services or other authorities, and increasing economic/staff resources.Endogenous examples are efforts to strengthen teachers' competence and give the school a certain profile (e.g., football, dance), and information/advertising campaigns to enhance schools' positions in a competitive market.

Analysis
In our analysis, we first address the respondents' descriptions of the prevalence and causes of school segregation in Swedish municipalities.Then we present our tripartite typology of local initiatives that seek to counter the negative effects of school segregation and the exogenous and endogenous drivers involved in each of them.Finally, we conclude by presenting our analysis of what national and local measures the respondents considered essential in reducing ethnic and socioeconomic segregation in the school system.

School segregation as a municipal problem
To contextualize the study, we begin by describing how the informants responded to the questions about if and how school segregation was a problem in their municipalities.Forty percent of the superintendents (79 of 197) stated that the schools in their municipality were ethnically and/or socioeconomic segregated, while 60% (118 of 197) said they were not.That school segregation does not exist in a majority of municipalities is to be expected.Informants in that category were usually from municipalities that were too small to have this kind of problem, that is, often the municipality had only one school.As one superintendent said, "The question is not relevant for our municipality when we only have one compulsory school; the chance of school segregation is therefore limited" (Superintendent 46).
If we turn to the 79 municipalities that noted a problem of segregated schools, the superintendents often described school segregation as leading to growing inequality and differences in school results between groups of students.This seems to be the most important driving force of countering the problem.One informant described this problem in the following words: We have a large difference between boys and girls and between students with parents that don't have secondary school education, compared to those that have secondary education or above.If we account for underlying factors and predictors, the expected result should be higher in this group [students from families with low socioeconomic status], especially in one of our lower-level secondary schools.We know that it looks just like this in many municipalities, which is not okay (Superintendent 60).
The majority of the superintendents in municipalities that struggled with school segregation (90%) also reported that they had local initiatives in place to mitigate the negative effects of segregation.

Three types of municipal initiatives to counteract school segregation
In this section, we conceptualize three types of municipal initiatives that seek to counter school segregation: reinforcement, dispersal, and merging.The analysis also lays out the exogenous and endogenous drivers involved in each of them and how they are used to create more heterogeneous student populations.In Figure 1, the three types of initiatives are visualized by highlighting their main characteristics, either transferal of resources or the relocation of students.

Initiatives of reinforcement
A common type of initiative described by superintendents was reinforcement, which entailed extra support to schools in structurally disadvantaged areas in order to improve the results and raise the quality of education in these schools, as well as attract a more heterogeneous student body.Our analysis shows that reinforcement was done in two ways: exogenously, by channeling external resources to the school, and endogenously, by enhancing resources that were already in place in the school.
In the data, the exogenous drivers were often the addition of funding and human resources or increased cooperation between schools and the community.Funds were transferred to schools either as a one-time, lump-sum payment or as increased funding over several years, often based on a local economic allocation model.One of the superintendents described this type of effort: We have an economic allocation model that accounts for socioeconomic factors such as parents' education, the share of newly arrived students and families on social allowance (Superintendent 121).
Another exogenous driver was to employ extra staff, such as resource teachers and teachers' assistants, extend the student health department, or increase homework assistance for students.One example of this was allocating two teachers to each student group: [We have] local efforts in our compulsory schools.Among other things we have a "two-teacher system", study supervisors and substitute teachers that are locally employed at the schools, increased student health, and local safety-and attendance teams (Superintendent 134).
Another exogenous effort that was noted in the survey was increasing the cooperation between the school and municipal services (social services, urban planning), external authorities (police), and civil associations (sport clubs, culture clubs).This represents a more holistic approach to addressing school segregation because the relationship between schools and society is emphasized.Often these efforts took the form of time-limited projects: The municipality has invested in housing in socioeconomically marginalized neighbourhoods (…) and [increased] cooperation between the municipal and the civil society, for example through sport clubs.But [there is] also increased cooperation between different administrations within the municipality, [as well as] organization and cooperation with governmental authorities (Superintendent 94).
The superintendents also described how they worked to reinforce the schools endogenously by strengthening teachers' competence, focusing on methods for educating newly arrived students, and supporting special pedagogical initiatives, such as multilingual teaching.
[We have] methods where we work with language development approaches, an area where we now have an extensive competence development in place for the teachers (Superintendent 134).
Often the superintendents described how they combined exogenous and endogenous drivers to reinforce the schools, but the most prevalent efforts were the exogenous ones, in particular, increasing funding to segregated schools.
In sum, reinforcement covers a broad spectrum of drivers that were intended to strengthen schools, staff, and students, as well as the interplay with internal (municipality services) or external actors (authorities, civil actors) within the local context.The municipalities often worked with several reinforcement drivers simultaneously, even though increased funding was the most common measure used to mitigate the negative effects of segregation.However, reinforcement also took the form of allocation of students and staff in order to create a more socioeconomically mixed student population in schools, as we discuss next.

Dispersal initiatives
It is also common for municipalities to change the mix of students in certain schools (see Arneback et al., 2021;Lund, 2021).One way of doing that is dispersing students to different schools in order to create more socioeconomically and ethnically mixed student bodies.The dominating drivers in this category were exogenous.For example, redrawing the lines of school catchment areas or busing students to schools other than those closest to them.Many of the superintendents were investigating these two measures as possible initiatives for the near future.A superintendent from a municipality that had instituted these measures, said, The municipality has had a large inflow of newly arrived students, and the board took the decision to bus these students to schools outside of the city centre.This was done to counteract the concentration of students with immigrant backgrounds in certain schools [and is] one way of creating more heterogeneous schools (Superintendent 29).
Another way to disperse students was to close a school, or a segment of it, and assign these students to other schools.This has been done in Malmö, Trollhättan, Linköping, and Örebro during the last decade (Arneback et al., 2021;Wigerfelt, 2010).One of the informants described a decision to disperse students: [We decided on] the closure of school X, classes 7-9, and (…) now in January about free transportation from neighbourhood X to school Y in order to get more students to choose school Y.We also took a decision to directly refer students in neighbourhood X to particular schools [in the municipality] (Superintendent 77).
In this case, students from neighborhood X had to choose a different school if they did not want to attend the school to which they had been assigned by the educational board.In the survey, some superintendents also mentioned that they were cooperating with the municipality-urban planning department to support mixed neighborhoods, for instance by building new flats or detached housing.The aim was to create more socially mixed neighborhoods and thereby create more socially heterogeneous schools.
There are also traces of endogenous drivers in the dataset.These drivers focused on increasing school choice, for example, by targeting parents with school information and offering free school transportation with the aim to increase choice.One informant in the survey described what had been done in their municipality: [We see] good results from giving all students a free bus card so they can choose a school without any [added] expenses, in combination with active work to increase the opportunities for students to choose a school within the municipality borders.We have students that move between different parts of the municipality, and we manage much better than expected to create equivalent results between the different schools (Superintendent 44).
In sum, dispersal initiatives were more direct than reinforcement initiatives, as they were meant to relocate students to other schools.Here, the exogenous drivers were more common than endogenous ones.The survey also revealed that dispersal often was directed at certain neighborhoods and schools, and students with a foreign background living in structurally marginalized areas.These students were often the target group for direct information about school choice or that had to relocate to other schools.

Merging initiatives
Merging means that the municipality closes two or more schools and creates a new one to ensure a more heterogeneous student composition at the new school (Kornhall & Bender, 2019;Wigerfelt, 2010).The incentive for merging schools is often economic since the new schools have the capacity for a higher enrollment.In those cases, a more heterogeneous student composition is a side effect.
A few municipalities had pursued this option or were planning to do so in the future.However, it was an uncommon strategy compared to reinforcement or dispersal.This type of initiative often relies on building new schools (an exogenous driver), and was envisioned by one of the informants in the following way: [We would] build a larger school in the city centre and in that way control the student population.Simultaneously, we would close some grades at the schools with the most homogenous student population, schools where parents have a low educational level, and a high degree of the students have a foreign background.We also plan to build schools between different neighbourhoods rather than in the middle of homogenous areas (Superintendent 108).There were no examples of endogenous drivers of merging in our data, although there were traces of endogenous merging efforts in the answers.For instance, schools in the same town but in different neighborhoods could be organized as a unit and students from different backgrounds interspersed within the unit's schools to create a more heterogeneous student population in each school.Respondents reasoned that this would be possible in smaller towns but would be difficult if the distance between schools was too great.This type of merging has been done in, for example, Denmark but with the aim to save schools in rural areas and create more economically solid units (Volby-Beuchert et al., 2016).It could also mean that schools have different pedagogical arrangements, for example, teachers divide their worktime between different schools or pedagogical learning between different schools (Ainscow & Howes, 2007).
In sum, merging schools to create heterogeneous student populations was an uncommon strategy in our data.In practice, it relies solely on exogenous drivers.It was most often mentioned in conjunction with building new schools, which municipalities often planned to build on the boundary of neighborhoods with different socioeconomic compositions in order to create more heterogeneous student populations.

What changes did the superintendents wish for?
In the survey, the superintendents were asked to reflect on what they thought needed to be done to reduce the negative effect of ethnic and socioeconomic segregation in the school system.The answers reveal two main themes: 1) Reducing residential segregation and 2) changing the national policy on school choice.The first theme, according to the informants, was especially important since the answers showed that the residential segregation was seen as the primary cause of school segregation.Especially in municipalities that have no or few independent schools (Andersson et al., 2012;Hansen & Gustafsson, 2016).One superintendent stated: In our municipality school segregation is a consequence of residential segregation, and school choice has a more limited significance.Efforts that limit residential segregation would also have an effect on school segregation (Superintendent 136).
When it came to the national policy on school choice, almost half of the superintendents in the study (46%) answered that it needed to be changed so that municipalities had more control and a larger mandate over their student population.Some thought that it would be best to remove choice completely in the interests of a better mix of students and others wanted to limit school choice.
You have to stop the "free school choice" and introduce national rules for diversity in the student population (Superintendent 116).

Another superintendent highlighted the need to
Remove the concept of "relative proximity" and give the municipalities greater opportunities to steer students between schools and neighbourhoods with the aim to create diverse student groups (Superintendent 10).
Connected to the issue of school choice, superintendents also described the need to remove the queuing system in independent schools.The answers reveal that superintendents thought that this was important since in their experience, independent schools were treated differently than municipal schools.However, they did not necessarily state in what way.
[We have to have] political discussions on a national level, with changes in the law around independent schools.The terms are not equal today since the independent schools have benefits that the municipal schools do not' (Superintendent 84).
Respondents noted how popular independent schools have long waiting lists, which means children have to be put on the list at an early age.The queuing system is seen as disadvantageous for students from families with a low educational level and socioeconomic status, and for newly arrived students who, for obvious reasons, cannot be placed in the queue in adequate time.Some of the informants argued that the queuing resulted in an uneven distribution of students based on educational and socioeconomic background and country of origin and that the queuing system in independent schools, therefore, needed to be changed: [We have to set in place] limitations regarding the free school choice, for example, through a lottery, to popular schools, instead of a wait list.The time criterion regarding the wait list discriminates students/families that, for example, are newly arrived and who have not been able to get on a wait list (Superintendent 142).
In sum, the superintendents called for two changes in educational policy to address the constraints they worked with in their efforts to counter school segregation.Most important according to them was to limit, or eliminate altogether, the choice of schools.This would put the independent and municipal schools on equal footing.However, superintendents also wanted municipalities to be given a stronger mandate and more opportunity to form socioeconomically and ethnically mixed student groups in their schools, something that they knew was lacking.While residential segregation was seen as giving rise to school segregation, most of the superintendents considered this harder to address and that the main responsibility for decreasing residential segregation lay elsewhere.

Summary and discussion
This section brings together the analysis and summarizes the three distinct initiatives used to counter school segregation, as well as constraints.The last part of this section describes the distinct gaps between different organizational levels in the educational system that need further attention.Finally, we end this section with some concluding remarks.

A typology of local initiatives to counter school segregation
Our study shows that school segregation is considered a prevalent problem in many Swedish municipalities.Forty percent of the superintendents answered that their schools were ethnically and socioeconomically segregated.In the other municipalities, a common answer was that school segregation was not an issue because of the low number of residents and schools.This indicates that a large proportion of mid-and large-sized municipalities in Sweden (around 50,000 people and more than 100,000 people, respectively) deal with these issues, but the answers also signal that some of the smaller municipalities (166 municipalities had less than 20,000 inhabitants in 2022) also deal with this problem.
Since the early 1990s, municipalities have been responsible for counteracting the effects of local school segregation.Our analysis reveals that municipalities use three main types of initiatives to counter school segregation: reinforcement, dispersal, and merging.All three rely on drivers that are used to create the desired outcome.We differentiated between exogenous drivers that originate from outside the schools and endogenous drivers that aim to strengthen the resources already in the schools.
Table 2 below summarizes our main results.
The most common initiative was to reinforce the schools, or in other words, to direct extra support to schools in structurally disadvantaged neighborhoods to raise their standard and thereby their results, and to attract more heterogeneous groups of students.This could be done in two ways, exogenously, by acquiring resources from the outside the school (e.g., money or staff), or endogenously, by enhancing resources that were already in place in the school (e.g., strengthen teacher competence).
Another common way to create heterogeneous schools was to disperse students.Municipalities closed a school or a part of a school, often in a structurally disadvantaged neighborhood, and assigned the affected students to other schools.Dispersal initiatives were initiated through exogenous drivers, such as redrawing school catchment areas, and endogenously by making school choice more accessible through advertising campaigns.
Merging was the most uncommon strategy mentioned by superintendents.Municipalities closed several schools or parts of schools and then consolidated the students in one school.This type of initiative relied on exogenous drivers, mainly building new schools in central locations or on the boundary of neighborhoods with differing socioeconomic compositions.The aim was to create a more heterogeneous student body.
Finally, the study highlights the constraints experienced by the respondents in their efforts to counter school segregation.A central constraint was residential segregation but also conditions created by national school policies.In the next part, we elaborate on these issues.

Minding the gaps
Another way to understand the results of the study is to examine the possibilities and constraints identified by the respondents in the study and to pay attention to what they did not address in their answers.Even if the municipalities are able to choose which initiative to implement, the survey points to several factors that hindered local anti-segregation initiatives.Some of these can be conceptualized as gaps, for instance, the varying degrees of the power and responsibility of actors at the different levels in the educational system.
Starting with the national and local levels, the empirical results of this study are in line with the many tensions that have been reported in research (Hennerdal et al., 2020;Pareliussen et al., 2019).The successive changes in national and international educational policies described above have been characterized as a shift from centralization to decentralization and back again to recentralization (Bergh & Wahlström, 2018;Granvik Saminathen et al., 2018;Grek, 2020).This means municipalities and schools are responsible for fulfilling national policy expectations, while they are simultaneously unable to control the many external factors that condition them.From our study, it can be concluded that the independent, for-profit school sector, juridical regulations on school choice, and emphasis on achievement in combination with local factors such as ethnic and socioeconomic residential segregation, clearly condition the local possibilities of countering the negative effects of school segregation.
Some of the informants in our study described how the control they have over student movement between schools is meaningless; in the end, local initiatives do not have an appreciable effect on school segregation because parents circumvent municipal directives.This is evident, for example, in Nyköping where the municipality closed several schools to merge the student bodies into a new one with the intention of creating a more heterogeneous mix of students.However, after a while, Swedish-born and well-educated parents began moving their children to independent schools in the area (Kornhall & Bender, 2019).If this trend continues, local initiatives like the ones in Nyköping (Kornhall & Bender, 2019), Örebro (Arneback et al., 2021), and Malmö (Wigerfelt, 2010) risk being fruitless efforts to reduce school segregation since superintendents are powerless in the face of independent schools with their queuing systems and the processes of residential segregation.
Turning to the local and school levels, our study reveals that many of the initiatives targeted structural and organizational conditions using mainly exogenous drivers with less attention being paid to how, for example, social and pedagogical processes can be initiated and supported in schools' daily pedagogical practices (Arneback et al., 2021).The analysis thus reveals that the municipalities have greater knowledge of and a tendency to turn to organizational processes, rather than focusing on and supporting social and pedagogical processes at the school level, which were uncommon in the data.This tendency seems fueled by the lack of cooperation between superintendents, headmasters, and teachers.Expressed in terms of exogenous and endogenous drivers, exogenous drivers dominated the responses, meaning that endogenous efforts need to be developed further.
A strength of the survey is that it reached many municipalities and by doing so produced a broad picture.On the other hand, the constraints of the survey meant that the respondents could not answer our questions in a fulsome manner.Further research using other methods is therefore needed to uncover more of the actual processes inherent in the initiatives, such as the above-mentioned social and pedagogical processes that take place within and between schools and at the municipal level.

Concluding remarks
This article contributes to knowledge regarding if and how municipalities in Sweden experience school segregation and how they work to counter its negative consequences.To grasp this complex phenomenon, we have presented a typology that offers important insights into the most common types of desegregation initiatives and the exogenous and endogenous drivers involved in each of them.In relation to the ongoing societal discussion on school segregation, our typology of local initiatives offers a language to use, and the gaps identified in the discussion show struggles that occur in Swedish municipalities that need to be further addressed.
An important insight is that municipalities tend to focus on organizational processes rather than supporting social and pedagogical processes.This implies that there is a need for further attention to be paid at the school level.Research shows that the outcome of local efforts to counter segregation are clearly affected by a range of social and pedagogical processes at the school level, from developing pedagogical competences (such as language-supportive, relational, intercultural, and emancipatory approaches) to learning how to work with socially and ethnically mixed student groups.This includes creating inclusive learning environments that acknowledge and make use of the plurality of students' experiences and that support exchanges and friendships between students with different backgrounds, qualifications and needs, which originate in the inequalities of life in a segregated society (Arneback et al., 2021).
Another important conclusion is that neither schools nor the central municipal educational administration can counteract school segregation on their own.The results of this article show that superintendents in municipalities are aware of how national regulations and policies hinder local efforts to counteract school segregation.In particular, the school choice policy is seen to work against municipal efforts to mix students with different socioeconomic backgrounds.The analysis thus shows that even though municipalities are responsible for providing equal education and enhancing social cohesion, they are clearly hindered by factors beyond their control.This points to the need for a more holistic view of school segregation, where the challenges and constraints identified in the discussion on gaps need to be addressed and resolved.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by the Swedish Research Council, Vetenskapsrådet [grant no 2020-04294].

Figure 1 .
Figure 1.The three types of initiatives.
2. Describe the main work that has been done to counteract school segregation at the compulsory level in the previous few years.Open answer 143 3. Describe (name) the work that you are planning to do in the future.Open answer 113 4. What do you think must be done to counter the negative effects of school segregation at the compulsory level?

Table 2 .
Main results of this study.
. Creation of a new school