Trajectories of Newly Arrived Migrants in the Swedish Introduction Program

In a 2010 policy reform, the centre-right Swedish government attempted to bring the area of integration policy into closer alignment with labour market policy for newly arrived migrants. In this paper, empirical data gathered from 181 individual case files from the Public Employment Service (PES) is used to analyse the different trajectories of migrants. I construct a typology of five different sub-groups to explore how migrants manage the different challenges of the policy program. The main contribution of the article is to highlight the individual differences among the group and how various strategies are formed within the same policy framework.


Introduction
Active labour market policy has been an important instrument in Sweden for combining economic growth and the inclusion of different groups in the labour market (e.g., Dahlstedt, 2009;Lundqvist, 2015;Montan, 1988;Nycander, 2008), in a sense foreshadowing more recent Anglo-American and European trends of workfare (labour market policy with rules to receive benefits) or activation (Brodkin & Larsen, 2013). In contemporary European welfare states, social investment (investing in people and giving support) has been put forward as a policy initiative aiming to include different groups into the labour market (de la Porte & Jacobsson, 2012;Kvist, 2015). One of these targeted groups is newly arrived migrants, where labour market and integration policy in the Nordic countries have often been combined through specific introduction programs. These programs aim to facilitate both labour market inclusion and more general integration ambitions; this ambition is expressed, for example, through civic education courses (Borevi, Jensen, & Mouritsen, 2017;Breidahl, 2017;Breidahl & Fersch, 2018;Hagelund & Kavli 2009).
In Sweden, a new introduction program labelled the "establishment reform" for newly arrived migrants was introduced in 2010 following other changes in labour market policy discussed below. In this text, newly arrived migrants will be used, referring both to individuals with refugee status and accompanying family reunification migrants, as these categories are both included in the policy discussed in the article. Responsibility for migrant integration was shifted from local municipalities to the Public Employment Service (PES). These changes can be interpreted as marking a rhetorical shift regarding how the needs of

Layering processes in the institution of labour market policy
The importance of past traditions has been highlighted in historical institutionalism as these are often seen to structure policy initiatives and thereby shape policy by facilitating certain policy choices over others (Rothstein & Steinmo, 2002) and allowing "policy feedbacks" to influence future policy choices (Béland, 2010). The incorporation of new groups into an institution can be one example of how institutions evolve incrementally in order to ensure stability (Thelen, 2004, p. 8). One example is what Thelen defined as "layering", namely the "grafting of new elements onto an otherwise stable framework" (2004, p. 35). In this article, I argue that labour market policy can be seen as an institution in its own right and that the establishment reform can be discussed as a case where new and old elements are combined in a process of layering. The combination of new and old elements in this institution is exemplified by the concepts social investment and workfare. These concepts have been chosen both as academic concepts, highlighting certain labour market trends that have been debated in the literature, but also as concepts that have guided policy choices and practical policy development (Ennerberg, 2017).
The two partly parallel trends within labour market policy -social investment and workfare-can more clearly emphasize different ways of understanding and constructing unemployment support. Changes in Swedish labour market policy can also be seen from a broader international perspective and have been discussed in terms of activation policy. Labour market policy designed to increase participation among individuals contains elements of, on the one hand, support to individuals, with the ambition to lead to empowerment and, on the other hand, controlling measures that can be seen as disciplining individuals (Jensen & Pfau-Effinger, 2005;Newman & Tonkens, 2011). This double sidedness can be seen historically through the emphasis of productive social policy or the intimate link between the development of the welfare state and social insurance primarily associated with workers (Morel, Palier, & Palme, 2012), even if different aspects have been emphasized more strongly during different periods (Edebalk, 2013).

Workfare
Workfare, or work-first, policy has traditionally been associated with Anglo-American labour market policy of the 1990s, where stricter rules were introduced in relation to labour market benefits (Brodkin & Larsen, 2013). These rules have included, for example, the requirement for individuals to participate in labour market measures in order keep their benefits. Workfare policies have also prioritized a fast entry into the labour market over finding a job within the individual's preferred vocational area or within the individual's educational or occupational background. Surveillance and sanctions have thus been used to change behaviours and attitudes among the unemployed. The individual's responsibility to find employment is here emphasized rather than the state's role in creating work opportunities (Brodkin & Larsen, 2013;Dostal, 2007;Lødemel & Moreira, 2014). Criticisms of workfare have primarily focused on the sanctions and controlling measures that have been used to steer individual behaviour (Newman, 2011). Some commentators have also argued that many of the activities that the individuals are required to participate in, such as CV courses or coaching activities, can be experienced as having limited or no meaning for individuals, and that they do not necessarily lead to increased labour market participation (Dahlstedt, 2009;Dostal, 2007). On the other hand, these controlling measures can be seen as necessary for legitimizing unemployment insurance in the eyes of employed groups, and they can also contribute to increasing individuals' efforts to find work (Cordula, 2013;Hohmeyer & Wolff, 2018).

Social investment
A different perspective that has gained much attention in recent years is social investment policy. The overarching goal here is to invest in individuals' education with the aim of strengthening their skills, thereby not only facilitating labour market entry but also providing support in times of unemployment (Kvist, 2015). In comparison to workfare measures, the state's role is thus emphasized, in particular in terms of providing labour market education, but also in terms of creating a labour market with "good jobs" for individuals (Bonoli, 2012;Morel et al., 2012). In a sense, social investment can be seen as a "win-win" policy: society supposedly gains by returns on the investment in terms of economic growth and in producing individuals with the right skills and education that are able to actively participate in society, while the individual benefits from being able to participate in relevant educational opportunities, and thereafter taking up more qualified work. Certain studies emphasize the importance of providing different support that can be adjusted to the individual's needs from a life-course perspective (Hemerijck, 2015;Kvist, 2015). Individuals' transitions between work or temporary periods outside of the labour market are supposed to be managed with support from social policy or unemployment policy, labelled by Hemerijck as "flow, stock and buffer", covering different phases of their needs, where, for example, buffer policies are used to support individuals through periods of unemployment (ibid). This broader interpretation of social investment could thus be seen as a more traditional social policy with its focus on economic and social support for the unemployed.
While the social investment perspective focuses on the role of the state to a greater extent than workfare measures, the perspective has also been criticized for emphasizing educational measures rather than policies that could reduce class and income differences through redistribution (Pintelon, Cantillon, Vand den Bosch, & Whelan, 2013). Despite the increased focus on women's participation in the workforce in the perspective, social investment has also been criticized from a gender perspective. For example, some authors argue that paid work is promoted at the expense of unpaid care work, which is not taken into account or is devalued (Jenson, 2009;Saraceno, 2015).

The Establishment Reform and the Case of Sweden
The exclusion of certain groups from the labour market was seen as an important policy issue for the centreright Alliance government that came to power in Sweden in 2006. To include these groups in the labour market, several changes intended to "make work pay" -such as changing rules regarding unemployment insurance, income tax levels, and introducing various activation programs targeted at different groupswere implemented (Kjellberg, 2006;Lindvert, 2015).
In 2006, municipalities were responsible for implementing integration policies for newly arrived migrants through locally organized introduction programs. The introduction programs were required by law to cover a range of individually adjusted labour market activities and the introductory course in Swedish for migrants (SFI). However, these introduction programs were criticized for failing to provide equal opportunities for migrants, depending on how municipalities chose to organize the programs, and for limited contacts with the PES. Moreover, men and women in the introduction program were seen as being treated differently, with men gaining more access to labour market activities than women. (Integrationsverket, 2004(Integrationsverket, , 2005(Integrationsverket, , 2007SOU, 2008).
Following a governmental white paper, a new reform named the "establishment reform" was introduced in 2010 with the explicit aim to shorten the time taken by newly arrived migrants to enter the labour market (Prop. 2009/10:60). The political rhetoric surrounding the reform portrayed the changes as a "paradigm shift" in Swedish integration policy (exemplified by the change of the policy area from integration to establishment), where previously "passive" and "caring" policies were to be replaced by an active policy focusing on work (Ennerberg, 2017;Larsson, 2015). Individual freedom of choice and active participation in the labour market were labelled as important components, not only to increase participation in work, but also to promote individual empowerment.
Through these changes in 2010, the state authority Swedish PES was given the main responsibility for the group newly arrived migrants. The shift of responsibility to the PES was partly seen as a way to prioritise employment in the integration process. Municipalities remained responsible for Swedish courses and were also given the task of providing civic education courses as one of the measures in the new introduction program (SOU, 2008).
The two-year-long introduction program was comprised of support from the PES through full-time activities that were to be determined in an individual action plan between each participant and their employment officer. These activities were supposed to build upon individuals' previous experiences and future labour market goals. The individual action plan usually included Swedish studies in combination with other measures, such as work experience, subsidized employment, on a job training or labour market education. The PES also used "complementary actors" to provide different labour market activities targeted to this group, including labour market training programmes and basic upskilling courses. This measure was eventually discontinued due to serious problems with the establishment guidance system (Riksrevisionen, 2014).
Through the measures, the importance of paid work was emphasised -with a focus on full-time activities, possibilities to gain different forms of work experience, and job subsidies, all of which were regarded as important aspects of the introduction program. The monetary benefit to participants (establishment benefit) was also constructed with a focus on financial incentives, where paid work and establishment benefit could be combined during a period of time, and where absence from activities led to benefit reductions (Prop. 2009/10:60, p. 105). The possibility to combine benefits with paid work was discontinued in 2016.
At the time this study was conducted (2011-2013), individuals within the introductory program who were unable to perform full-time could participate at 25%, 50% or 75% levels. PES could also include preparatory activities focusing on, for example, health or social support in the individual action plan. For participants with more serious health problems, the PES used specialists to evaluate needs related to certain health issues, with a focus on finding rehabilitation measures that could facilitate work (SFS 2010:409).

Aim and research question
This article sets out to trace changes in Swedish labour market policy from a perspective inspired by historical institutionalism and the processes of layering (Thelen, 2004). The main aim is to explore how continuity, rather than breaks in policy, shape the policy outcomes for a group of migrants. The article is guided by the following research questions: (1) How can newly arrived migrants' trajectories in the introduction program be constructed? (2) How do the changes in labour market policy influence migrants in practice? The aim is to provide a snap shot of the group often referred to in policy documents as "newly arrived migrants". A more nuanced understanding of the differences within this group may be used for better targeted policy initiatives.

Methodology
This case study is based on a qualitative documentary analysis based on data covering the years 2011-2013 from the PES. This material consists of individual action plans, notes from the Public Employment Service for individual case files and monthly reports from private actors involved in the introduction program. In the initial study, interviews were conducted with PES employment officers and establishment guides. These interviews are not directly referred to in this paper, but the material and the analysis of these may, to a certain extent, have influenced the current analysis (see Ennerberg, 2017Ennerberg, , 2020. The documentary data consists of 181 cases with three different types of documents: individual establishment plans, monthly reports written by establishment guides, and a printout from the PES's internal documentation system for each individual. The internal documentation from the PES included information such as the individual's work history, planned activities, contacts with employers, external providers and establishment guides and reduction of benefits due to different reasons. Through these cases, I was able to follow individuals throughout their establishment period, for example by gaining information about which activities they were participating in. The individual case files vary in length, from approximately 5 pages (if, for example, the individual immediately finds work or for other reasons exits the introduction program) to around 120 pages (if, for example, the PES has formulated extensive notes about the individual's time in the program and/or the individual frequently changes activities and/or establishment guides). Due to the sensitive and personal information available in these documents, ethical approval was applied for and granted by the Swedish Ethical Review Board, and passive consent was requested from the participants through a written letter. The selection of cases was made by staff at the PES, where, in three different offices, case files were requested of the first 40 individuals who had exited the introduction program during three or four consecutive months in 2013. Reasons for leaving were most commonly that the individuals had completed the 24-month introduction period, had found work, had retired or had gone on parental leave. All in all, 197 cases were selected. A small number of cases were removed for those participants who asked to be removed from the data set. The cases where participants exited the program after only one or two months, and where there was insufficient information in the case files for these to be analysed, were also removed from the database, thus leaving the 181 files that comprised the data.
Based on reading and interpretation of the material, the cases were used to construct five different ideal types of trajectories in the establishment period. Typologies in the social sciences have, since Weber, been seen as a way not to fully describe reality, but as an analytical construction where certain elements are accentuated or clarified (Weber, 2017(Weber, [1949)). Typologies can also be used to make an extensive material more comprehensible (McKinney, 1969) and to analyse sensitive information without revealing data about individuals (Widerberg, 2002). The construction of ideal types necessitates a differentiation between the different categories, where certain elements are accentuated more than others. In this sense, some similarities between the categories, as well as some nuances, may be neglected to more clearly construct the categories as different from each other (Esaiasson, Gilljam, Oscarsson, & Wangnerud, 2012).
The ideal types were constructed through initial reading, where certain basic information about the individual cases -such as gender, age, municipality and activities individuals participated in -were registered and coded in a database. After this coding, the documents were read through several times, after which a shorter summary of each individual's time in the establishment was written down. These preliminary notes made certain recurring patterns visible in terms of how individuals went through the introduction program. In the next step, longer summaries that tried to capture these different routes were written down, in essence summarizing several cases and their trajectories. After constructing five summaries, I returned to the individual cases to consider how these fitted the constructed ideal types. After this re-reading, changes to the ideal types were made, and individual cases were sorted to see whether they all could "fit" into one of the constructed ideal types. It was noted that certain individual cases could possibly fit more than one category, but category that was seen as most appropriate after reading the case files more closely was chosen. Through this process, aim was to find a way to describe the process individuals went through as part of the introduction program and to capture the central aspects that I have seen as signifying the different trajectories.
A weakness with the chosen method was that the individual cases that contain more extensive descriptions and notes to a higher extent may have influenced my reading and construction of the types. Another weakness is that the material mainly builds upon secondary data and notes from employment officers at the PES and establishment guides. These actors can be seen as having a certain stake in the process, rather than being neutral commentators. Thus, interviews with participants would most likely give a different picture of the process. Consequently, the notes may paint a more positive picture of the actors' own organization. The material also covers both the activities that participants take part of and certain background information and different strategies and reactions from the individuals (based on comments from others). The material is a mix of, on the one hand, certain background information and, on the other hand, both strategies and participation in the introduction program.

Results
By summarising the different ideal types developed from material, I will show the heterogeneity of this group and point to some particular difficulties that exemplify the needs and capabilities of the group as seen in some of the cases. After presenting the trajectories of migrants, I discuss how the ideal types are represented in the material, as well as the consequences and support provided through the establishment program.

Swedish learner
This first ideal type represents individuals who primarily use the introduction period of two years to study. Swedish language learning was important for these individuals to pursue further education in Sweden. These individuals often progressed rapidly through the Swedish courses. Others in this group who already had a higher education degree from their home countries wanted to validate their education to, thereafter, be able to work within their previous line of work, once they had the necessary Swedish language skills. The individuals in the Swedish learner ideal type were less interested in other work-related activities. In other cases, they took up jobs offered through private establishment guides or employment officers, or they found part-time work on their own, which they combined with their studies.
Sometimes they managed to complete Swedish courses beyond the Swedish for Immigrants level, as well as other courses necessary for progressing with higher education. If they already had higher education degrees from their home countries, a validation of their skills combined with Swedish language skills was sometimes enough to enter work. If the degrees were difficult to transfer to the Swedish labour market, a shorter course of education in a similar area was sometimes seen as more desirable.

Frustrated jobseekers
This second ideal type describes individuals with a clear goal of finding employment, but who often struggled to enter the labour market and fulfil this goal. Out of the five ideal types, this was the most diverse group studied, both in terms of background and in terms of how they managed the establishment process.
Many of the individuals in this group were either in the younger age group or older individuals who had previously worked in their home countries. Their main aim was to find work as quickly as possible and, rather than having a clear professional ambition, they were interested in finding any job. Individuals of this ideal type often combined Swedish for Immigrants courses with vocational Swedish or skills portfolio courses, but they were more interested in work placements. In some cases, these individuals struggled with the language and/or workplace norms. Language difficulties could lead to employers finding that they had to spend too much time mentoring these individuals, or that mistakes were often made. Those who had undergone several work placements that had not led to permanent work often expressed frustration and wavering motivation -sometimes feeling used by employers as free labour.
The lack of success in finding stable employment often caused individuals in this ideal type to question the possibilities of finding work, leading them to redirect their efforts into labour market training programmes or to study Swedish for Immigrants full time with the hope of finding a new job in the near future.

Work settlers
The third ideal type comprises individuals who, to the greatest extent, found work in the Swedish labour market, either through subsidized types of work available for individuals in the establishment programme or through regular work. By the end of the introduction period, these individuals had found employment, either combined with different labour market programmes or they were engaged in studies. Their way through the introduction measures, however, varied. Such individuals often lived in one of the larger cities and were usually motivated to find work at the start of the introduction period, while also studying Swedish for Immigrants. They often took the initiative to find jobs, sometimes through previous contacts or a local network and in the service-branch in areas such as cleaning, retail or restaurants.
These individuals were often able to find work placements that may lead to certain subsidised employment. In many cases, however, employers were reluctant to hire them on a more permanent basis or in a less subsidized form of employment. Nonetheless, at the end of their introduction period, and after trying different types of jobs and activities, individuals of this type often received increased help with job matching from the PES, and could thereby find more stable employment.

Establishment strugglers
The fourth ideal type comprises individuals who struggled to participate in establishment activities and who did not make clear progress during the two-year period. Establishment strugglers developed less connection with the labour market through the introduction measures, and they struggled to find a suitable path throughout the establishment period. These individuals often lacked clear goals in terms of labour market participation.
Some individuals of this ideal type had a high rate of absence from activities or, according to the external providers or Swedish teachers, appeared unengaged. In some cases, individuals in this type travelled for a period of time without notifying the PES, often to see family members abroad and to bring them to Sweden. In these cases, absence could further break the continuity of the introduction period. Many of these individuals participated in shorter work placements that ended due to issues with timekeeping or other misunderstandings between individuals and employers.
During some periods, these individuals articulated clear labour market goals that required further studies and chose to focus solely on Swedish for Immigrants. However, progress was often slow; and when employment officers discussed career choices and explained the nature of the work in further detail or the study path towards a professional qualification, these individuals often changed focus to a different type of job with less demanding qualifications. Such individuals would often not make clear progress after the introduction period.

Hindered establishers
The fifth ideal type comprises individuals who struggled to participate in the establishment measures due to health and/or social issues. These individuals generally participated in Swedish studies throughout the period, although the participation rate varied depending on their well-being during the two years. Some could participate in adjusted activities with a particular health focus that was offered locally. In other cases, they participated full time in the programme through the work preparatory activities that employment officers were able to include in individual plans. Employment officers followed up on the health and social status of these individuals through PES experts or outside professionals, such as doctors, in order to review the situation.
Many individuals suffered from health problems or related social issues that were a direct consequence of traumatic experiences of war and conflict. In these cases, they often needed to combine Swedish studies to a manageable extent with other activities directed at increasing well-being in the long-term. Handling social issues related to housing, divorce, schooling, etc. was particularly challenging and required adjusted activities at different times in the introduction period. Employment officers sometimes needed to establish contact with other authorities.
Individuals could find their way through the introduction measures when social issues seemed less pressing. Those with specific health problems that qualified for additional support through other PES measures could also find employment at the end of the period. However, those with uncertain medical conditions and low confidence in finding a job would often not find clear success at the end of the establishment period and thus focused primarily on the Swedish for Immigrants course.

Distribution of ideal types
It is important to take into account that this material cannot be seen as representative for the participants in the introduction program. Nevertheless, it can be interesting to consider how the five ideal types are distributed. As we can see from Table 1, work settlers make up the largest group in the material, followed by frustrated jobseekers.
As we can see from Table 2, women are considerably more frequently found in the categories Swedish learners and Hindered establishers. Among men 36% can be found in the work settlers category, compared with 15% of women.  For older age groups, 50% are found in the category hindered establishers, whereas 32% of the youngest age category belong to the frustrated jobseeker type (see Table 3).

Discussion
The ideal types that can be seen as most successful in terms of participating in work or studies seem to have been able to utilize the opportunities to independently choose those measures that fit their personal goals. For example, the Swedish learner often used the opportunity to study Swedish during their establishment period to later on be able to participate in the labour market in an occupation they saw as desirable. Others within this category also used the opportunity to be able to combine studies with paid work. The type of unskilled work that they participated in was often used as an extra income, but with the goal to later enter a different profession. For this group, it is possible to see the more positive and empowering aspects of activation policy (Jensen & Pfau-Effinger, 2005) including the workfare emphasis on economic incentives that are normally seen as controlling (Dostal, 2007). A different interpretation however, is that this group, which often had some form of relatively high educational level, has been prioritized by employers, thus leading to their relative success in terms of work opportunities. The added institutional elements such as sanctions and incentives have been incorporated into labour market policy and on the practice level can be seen as having been utilised by some individuals in line with the policy intentions.
The work settlers consisted of participants who, to a relatively large extent, found work during their establishment period, often through subsidized employment. These jobs were found either through their own networks or with help from the PES. Thus this group was similar to the Swedish learners in the sense that they were able to use the available measures within the introduction program. The strategy can possibly be seen as riskier, as subsidized employment might not lead to a permanent position. The added institutional elements of targeted subsidized employment here seem to serve as a bridge towards the labour market. Thus, to a certain extent, these elements build upon more traditional matching strategies that historically have been important for Swedish active labour market policy (Nycander, 2008).
One possible interpretation is that these two categories may have benefitted from the more intense and earlier contacts with the PES compared to the municipality introduction measures. For these two ideal types, the combination of positive and negative incentives may have worked to their benefit, in providing additional support as well as independence to navigate the program, in a sense symbolising the ideal "active" and "responsible" individuals, i.e. the more enabling aspects of activation policy (Newman and Tonkens, 2011).
For the Swedish learners with clear study ambitions, the time in the introduction program can be seen as an expression of the more ambitious social investment perspective, where a group outside of the labour market gets access to better career opportunities through further education. Both Swedish studies and studies at upper-secondary school level are used by this category as a goal-oriented strategy to advance to work, which without some of the support through the program would not necessarily be available for the group. For the ideal type frustrated jobseekers, the use of labour market education seems to offer a similar opportunity to be able to enter the labour market through participation in a specific measure, in the form of a new alternative career opportunity. The use of labour market education is not necessarily connected to an individual career goal, but more to a desire to find a possible career path, with the goal of economic self-sufficiency. Labour market education, on the occasions it does lead to work, can thus be seen as a potentially efficient social investment tool, which has also historically enabled the unemployed to (re-)enter the labour market (Bonoli, 2012).
The difficulties of workfare strategies can be seen through the other ideal types such as the establishment strugglers and the frustrated jobseeker. These categories are similar to one another in that during the introduction program they did not have great success in the regular labour market, even though the frustrated jobseekers to a greater extent found more opportunities to try to enter the labour market, for example, through work experience placements. Participation in these placements in those cases when they did not lead to paid work and were experienced by the individuals as negative can be connected to those aspects of workfare that emphasise control and discipline. As these categories do not immediately find a suitable path, the workfare elements of the program are seen most clearly -as employment officers at the PES in certain cases can use economic sanctions or promote different educational measures to ensure full-time participation. Workfare measures can thus here be used in terms of steering individuals towards certain activities, but they are more rarely used to direct individuals to a certain job. The main characteristic for establishment strugglers, in my interpretation, was the difficulty to find a suitable route through the introduction program. This group, with many younger individuals, might to a larger extent have needed more information and guidance regarding possible routes into work. Instead, the failure of this group to find an appropriate plan led to more tensions compared to the other categories, and experienced the clearest elements of workfare, where employment officers more actively tried to steer individuals to participate in certain measures, despite some individuals openly questioning the link to the labour market or the relevance for them personally. As noted above, the program also contained different more basic labour market guidance tools. Courses that focused on job-seeking strategies, or CV writing skills, rather than on specific education measures were criticized by some participants in the material, above all establishment strugglers, who did not always see any clear benefit in participating in these courses. Furthermore, certain Swedish learners did opt out of these courses, preferring increased hours in SFI courses, which were needed to achieve formal Swedish language skills necessary for further studies. For this group, success in their studies enabled this category to pick and choose which courses were necessary in terms of long-term career planning, and individuals in this group often had no problems in terms of resistance from employment officers. For establishment strugglers who, as mentioned above, did not have a similarly specific career plan, resistance towards these courses sometimes led instead to conflicts, particularly in relation to the demand of full-time activities. In this sense, participants with more successful establishment trajectories were more able to ascertain their autonomy than individuals who could not demonstrate such a clear career route (Newman & Tonkens, 2011).
Swedish studies can be interpreted as important from the perspective of a broader integration goal of societal inclusion, where all groups, even those for some reason unable to participate in full-time work, are seen as in need of Swedish skills to participate in Swedish society. Within the introduction program, Swedish studies was the one activity which all groups were expected to be able to participate in. Rather than an ambitious social investment measure, participation in Swedish studies can, for these participants who struggle to take part in other activities, be seen as an integration measure. Nevertheless, if social investment ambitions are to be realized, basic language training must be seen as a first step on a longer path towards more advanced language training, potentially followed by other educational initiatives. Separating social investment ambitions from other political ambitions in this sense becomes more difficult in practice.
Other elements of the introduction period for hindered establishers can be seen as characterized by a time-consuming establishment process, with a focus on evaluating the participants' ability to take part in different activities and on solving different problems. For this group, with complex problems that required more acute solutions, the social investment opportunities, could thus be seen as unavailable within the time frame of the introduction program.
By focusing on the different elements of workfare and social investment, this study has shown how the introduction program in practice incorporates new and old elements through what can be seen as "layering". The combination of "new" policy elements, such as the strengthened work focus, with continuing practices focusing on social support (that were seen by policy makers as impeding quick labour market entry) emphasises the institutional stability on the ground. The social support practices could, for example, override the new elements for the fifth category "hindered establishers". The different ambitions of integration policy and active labour market policy as set out in the policy are thus revealed also in practice, where certain "gaps" or dilemmas in terms of how to combine new and old policy elements are managed by case-workers. Through some ideal types, the ideas of a stronger work line are clearly visible, whereas in other categories, broader and more diverse concerns for the individual's integration process serve to continue a more "caring" policy line even after the new policy reform had been introduced.

Conclusion
This article has shown that the changes introduced in Swedish integration policy in 2010 contained elements of both workfare and social investment, but that the "carrots and sticks" in the policy framework also influenced the group of migrants to a varying extent. By constructing five ideal types I have showed how migrants' trajectories in an introduction program in Sweden can be constructed. The main contribution of these categories is to provide a more heterogeneous picture of this group of newly arrived migrants and their different trajectories through the first years in a new national labour market.
I argue that rather than exemplifying a clear shift towards stronger workfare for migrants, the early years of the new introduction program was thus instead built on a mix of different elements and ambitions, many of these building on previous integration policy, but also aligned with other labour market changes, emphasising policy stability rather than a clear break. Moreover, due to the different needs and backgrounds among the participants, it also seems as if social investment concerns have in many ways been incorporated into this labour market programme, which acts to limit the extent to which workfare policy goals can be carried out. For activation policy, more generally, this study contributes by shedding light on the manner in which certain individuals are, to a greater extent than others, able to act independently within the boundaries of existing frameworks and policy constraints.