‘Is My Volunteer Job Not Real Work?’ The Experiences of Migrant Women with Finding Employment Through Volunteer Work

We examine whether, and under which conditions, volunteering contributes to migrant integration. We identify two main goals of workfare volunteering—empowerment and employability—which build on two distinct images of the ideal citizen: the empowered citizen and the worker-citizen. Life story interviews were held with 46 first- and second-generation migrant women from Turkey, Morocco and Suriname living in the Netherlands. We found that volunteering contributes to employability and empowerment. However, for two mutually reinforcing reasons it eventually disempowers. Firstly, volunteering hardly ever results in paid employment because employers do not recognize volunteering as real work experience. Secondly, the focus on paid employment as ultimate form of integration misrecognizes migrant women as active citizens, which often results in disempowerment. Our findings show that the double policy goals of workfare volunteering require different conditions, and as such striving for both simultaneously often results in failing to achieve the set goals.RésuméNous avons examiné si, et dans quelles conditions, le bénévolat contribue à l’intégration des migrants. Nous identifions deux principaux objectifs du bénévolat dans le contexte de l’aide sociale, habilitation et aptitude à l’emploi, qui s’inspirent de deux images distinctes du citoyen idéal : le citoyen habilité et le travailleur. Des entrevues ont été réalisées auprès de 46 migrantes de première et seconde générations de la Turquie, du Maroc et du Suriname vivant aux Pays-Bas. Nous avons découvert que le bénévolat contribue à l’aptitude à l’emploi et à l’habilitation. Pour deux raisons mutuellement complémentaires, il est toutefois éventuellement paralysant. En premier lieu, le bénévolat mène rarement à l’obtention d’un emploi rémunéré, car les employeurs ne considèrent pas qu’il s’agit d’une véritable expérience de travail. En second lieu, l’accent mis sur l’accès à l’emploi rémunéré comme forme ultime d’intégration reconnaît injustement les migrantes en tant que citoyennes actives, entraînant souvent la disparition du sentiment d’habilitation. Nos résultats démontrent que les deux objectifs politiques du bénévolat dans le contexte de l’aide sociale requièrent différentes conditions, et que les efforts d’atteinte simultanée de ces derniers ne résultent fréquemment pas en l’ascension à l’un et l’autre.ZusammenfassungWir untersuchen, ob und unter welchen Bedingungen eine ehrenamtliche Tätigkeit zur Integration von Migranten beiträgt. Dazu identifizieren wir zwei Hauptziele der gemeinnützigen Arbeit im Rahmen des Workfare - Empowerment und Arbeitsmarktfähigkeit - welche auf zwei unterschiedliche Vorstellungen eines idealen Bürgers aufbauen: der befähigte Bürger und der arbeitende Bürger. 46 Migrantinnen erster und zweiter Generation aus der Türkei, Marokko und Suriname, die in den Niederlande leben, wurden über ihre Lebensgeschichte befragt. Wir stellten fest, dass ehrenamtliche Tätigkeiten zur Arbeitsmarktfähigkeit beitragen und zu Empowerment verhelfen. Allerdings wirkt sich die ehrenamtliche Arbeit aus zwei gleichermaßen bekräftigenden Gründen letztendlich entmachtend aus. Erstens führt eine ehrenamtliche Tätigkeit fast niemals zu einer bezahlten Arbeit, weil Arbeitgeber ehrenamtliche Tätigkeiten nicht als wirkliche Arbeitserfahrung anerkennen. Zweitens werden Migrantinnen als aktive Bürgerinnen verkannt, wenn der Fokus auf eine bezahlte Beschäftigung als ultimative Form der Integration liegt; dies führt oftmals zu einer Entmachtung. Unsere Ergebnisse zeigen, dass die beiden politischen Ziele der gemeinnützigen Arbeit im Rahmen des Workfare unterschiedliche Bedingungen erfordern und somit die gleichzeitige Verfolgung beider Ziele häufig dazu führt, dass die gesetzten Ziele nicht erreicht werden.ResumenExaminamos si el voluntariado contribuye a la integración de inmigrantes, y bajo qué condiciones. Identificamos dos metas principales del voluntariado de “workfare” (obligación de trabajar para recibir subsidios) - empoderamiento y empleabilidad - que se basan en dos imágenes diferentes del ciudadano ideal: el ciudadano empoderado y el ciudadano trabajador. Se realizaron entrevistas biográficas a 46 mujeres inmigrantes de primera y segunda generación de Turquía, Marruecos y Surinam que viven en los Países Bajos. Encontramos que el voluntariado contribuye a la empleabilidad y al empoderamiento. Sin embargo, por dos motivos que se refuerzan mutuamente les desempodera eventualmente. En primer lugar, el voluntariado raras veces resulta en empleo pagado porque los empleadores no reconocen el voluntariado como una experiencia de trabajo real. En segundo lugar, el foco de atención en el empleo pagado como forma última de integración reconoce erróneamente a las mujeres inmigrantes como ciudadanas activas, lo que da lugar a menudo al desempoderamiento. Nuestros hallazgos muestran que las dos metas de la política de voluntariado de “workfare” requieren diferentes condiciones, y como tales, esforzarse por ambas simultáneamente a menudo resulta en no lograr las metas establecidas.Chinese我们查看了志愿是否及在哪些条件下为移民融入作出贡献。我们确定了两个主要的工作福利志愿目标 - 赋权和就业能力 - 这构建理想居民的两个独特形象:赋权公民和员工公民。采访了46位生活在荷兰,来自土耳其、摩洛哥和苏里南的第一代和第二代女性移民的生活故事。我们发现,志愿活动对赋权和就业能力作出了贡献。然而,出于两个共同增强的原因,这最终会失去影响力。首先,志愿活动很少会导致有薪雇用,因为雇主不认可志愿活动是真正的工作经验。其次,将有薪雇用作为最终融合形式的重心会错误地将移民女性确定为活跃公民,这通常导致失去影响力。我们的调查结果表明,工作福利志愿的双政策目标需要不同的条件,因为同时努力通常会导致无法实现设定的目标。Japaneseどのような状況下であれ、ボランティア活動が移民の統合に貢献するか否か調査する。勤労福祉制度のボランティア活動の2つの目的であるエンパワーメントと雇用される能力を明らかにする。ボランティア活動の勤労福祉制度とは、権限を与えられた市民と労働市民という理想的な市民に2つの異なるイメージを構築する。オランダに住むトルコ、モロッコ、スリナムの第一世代および第二世代の移民女性46名の体験談を調査した。ボランティア活動は雇用される能力とエンパワーメントに貢献することがわかった。しかしながら、2つの相互的な強化は奪われる。まず、ボランティア活動は、雇用者がボランティア活動を現実の職業体験として認めていないため、有償雇用にはならない。第二に、統合の究極の形として、移民女性を活動的な市民として見ていないために権限が与えられず、雇用に注意が払われない。結果から、勤労者のボランティア活動方針の目的は異なる状況を必要としているが、両方を目指しても設定目標を達成できないことがわかった。Arabicنحن ندرس إذا كان العمل التطوعي يساهم في إدماج المهاجرين، وفي ظل هذه الظروف، نحن نحدد هدفين رئيسيين للعمل التطوعي في التدريب على العمل - التمكين وقابلية التشغيل - التي تعتمد على صورتين متميزتين للمواطن المثالي: المواطن المخول والعامل والمواطن. أجريت مقابلات على مستوى الحياة مع 46 امرأة مهاجرة من الجيل الأول والجيل الثاني من تركيا والمغرب وسورينام الذين يعيشون في هولندا. وجدنا أن التطوع يسهم في التوظيف والتمكين. مع ذلك، لسببين يعزز كل منهما الآخر، فإنه ينفي في نهاية المطاف. أولا، العمل التطوعي لا يؤدي أبدا” إلى عمل مدفوع الأجر لأن أصحاب العمل لا يعترفون بالعمل التطوعي كخبرة عمل حقيقية. ثانيا”، إن التركيز على العمل المدفوع الأجر بوصفه شكل نهائي من أشكال التكامل يسيء فهم النساء المهاجرات بوصفهن مواطنات ناشطات، مما يؤدي في كثير من الأحيان إلى عدم القدرة على العمل. تبين نتائجنا أن أهداف سياسة مزدوجة للعمل التطوعي في التدريب على العمل يتطلب ظروف مختلفة، وعلى هذا النحو والسعي لكل في وقت واحد في كثير من الأحيان يؤدي إلى فشل في تحقيق الأهداف المحددة.


Introduction
The Netherlands, just like many other welfare states with a considerable migrant population, has adopted several policies to stimulate migrant integration. Mandatory 'integration courses' (Asscher 2016), revoking residence permits in case of a lack of effort to complete the integration course, and signing a 'participation agreement' (Asscher 2015b) are some examples of recent measures. A policy goal that receives particular attention from the Dutch government is labor market integration. Labor market integration should reduce dependency on welfare benefits and contribute to self-reliance (Asscher 2015a).
Despite the aforementioned policies, there are still considerable differences in employment rates across migrant groups. Especially, migrant women seem to fall behind. The unemployment rates of women of Turkish (19%), Moroccan (16%) and Surinamese (15%) descent are disproportionately higher than the unemployment rate for native Dutch women (6%) and those of migrant men (Huijnk et al. 2014).
Volunteering is often perceived as a form of active citizenship that stimulates migrant integration into Dutch society and labor market integration more specifically. From 2004 on, the Dutch government has entrusted municipalities to implement 'workfare volunteering' (Kampen 2014)-a policy that requests welfare clients to volunteer in return for their welfare benefits-while at the same time incorporating volunteer work as an official part of migrant integration courses. 1 Even though in many Western welfare states volunteering is presented as a way of integrating in mainstream society, not many studies have looked into the effect of volunteering on migrant integration. The few studies have found that volunteering may fulfill its assumed role as a tool for migrant integration by offering a route to paid employment (Dudley 2007;Handy and Greenspan 2009). Based on these studies, we would expect (workfare) volunteering to be a promising way of improving the labor market prospects of Turkish, Moroccan and Surinamese women in the Netherlands. However, studies on workfare volunteering found that it also results in various negative outcomes that may affect labor market opportunities, such as feelings of disempowerment (Marston 2005), a loss of a sense of agency (Bessant 2000;Owen 1996) and feeling dependent (Andersen 2002).
In this study, we set out to examine whether, and under which conditions, volunteering contributes to integration of migrants in society at large and on the labor market more specifically. We develop our theoretical framework by identifying the existing policy goals and rationales behind workfare volunteering as a tool to promote integration and to relate these policy rationales to the models of citizenship from which they originate (Newman and Clarke 2009, p. 163). Subsequently, we discuss previous research on the consequences of workfare volunteering for welfare clients in general and the role of volunteering in migrant integration specifically. After discussing the methodology, we present the findings of our study regarding the extent to which the main goals of workfare volunteering policies-empowerment and employability-are reached, under which conditions these goals are reached, and which role social capital plays in reaching these goals. In the concluding paragraphs, we discuss the implications of our research and policy recommendations.

Mainstreaming Integration into Welfare Policy
Migrant integration is a contested term, with a shifting definition depending on the i.a. historical and political context (Favell 2001). Dutch policies on migrant integration have, for example, become increasingly demanding, encompassing and punitive, putting increasingly more pressure on the individual responsibility of migrants to assimilate in increasingly more aspects of life (Vasta 2007). Examples of more demanding and punitive policies are the new responsibility of migrants to pay for mandatory integration courses and withdrawing residence permits in case of failing the integration examination. Entzinger describes this policy development as raising the bar for those targeted by Dutch integration policies and critizes the Dutch government for ''changing the rules while the game is on'' (Entzinger 2006, p. 121). However, lately there has been a move away from specific migrant integration policies toward integration issues being mainstreamed into policies directed at the general population (Entzinger 2014). For example, policy measures like workfare volunteering are expected to also contribute to the integration of migrants.
While the increasingly demanding definition of migrant integration results in raising demands and expectations of migrants, mainstreaming seems to have resulted in a decreasing consideration of the specific obstacles that migrants face by equating migrants to the general population. Mainstreaming policy for migrant women on welfare might have different consequences going in roughly two directions. On the one hand, equating expectations from migrant women, who have the highest unemployment rates in Dutch society, to the expectations of welfare clients in general, seems rather ambitious. On the other hand, mainstreaming might also result in migrant women being treated as every other target group of interventions aimed at promoting active citizenship.
In order to investigate how these expectations play out in practice, we build on citizenship theory. Citizenship theory helps us to make sense of the citizenship ideals that dominate Dutch activation policy. Newman and Clarke (2009) point to the diversity of sites, forms and techniques that are at play in promoting active citizenship and show how several images of the ideal citizen 'swirl around' contemporary policy discourses that may be contradictory or bundled together in particular strategies. When looking more closely at workfare volunteering as a tool for migrant integration, it seems to be simultaneously based on the model of the 'empowered citizen' as well as the 'worker-citizen' (Newman and Clarke 2009).
The policy measure is in place since 2004, but gained momentum 5 years later. In October of 2009, two deputy ministers jointly sent a letter to the Dutch parliament, in which they formulate two goals of workfare volunteering: empowerment and employability. Both goals are presented as being dependent on an increase in social capital.
The first goal of workfare volunteering is empowerment (Kampen 2014), i.e., a passage from a state of helplessness to a state of better control of one's life, fate and surroundings (Rappaport 1985;Wieck 1983). According to the deputy ministers, 'people with a great distance to the labor market have often lost faith, no longer know their own strength, and lack the courage that is needed to find a job.' They go on to explain how volunteering helps people to regain self-confidence and control over their life by having their capacities confirmed by others.
The second goal of workfare volunteering is increasing welfare clients' employability, i.e., preparing people for reintegration into the labor market by teaching them skills and capabilities to find and perform in a paid job (Kampen 2014). In the same letter, the ministers claim volunteer work can 'contribute to active participation as part of reintegration and as a step toward the labor market, because volunteering prevents the loss of expertise of those who lost their jobs.' In addition, the deputy ministers claim volunteering 'keeps the work rate up to standard' and 'the unemployed can learn new skills' through volunteering.
Empowerment and employability may amplify or oppose each other. Amplify, because a boost in self-confidence and control over one's life may be expected to increase someone's employability. Oppose, because of the two conflicting images of the ideal citizen they represent. In the model of the empowered citizen, there is a strong focus on the needs and power of the citizen. Empowered citizens engage in decisions about their own welfare, and there is a strong focus on choice and voice. In the model of the worker-citizen, on the other hand, there is a stronger focus on obligations, responsibility, productivity and performance (Newman and Clarke 2009, p. 163). So the goals of workfare volunteering seem to be based on two contesting models of citizenship: a less demanding one focusing on citizen empowerment and a more demanding one focusing on obligations and responsibility. According to Cox (1998), activation policies that are based on such competing visions may 'operate sometimes at cross-purposes, and offer half-hearted alternatives'' (Cox 1998).

Outcomes of Workfare Volunteering
After having introduced our theoretical framework, we will now first discuss the outcomes of volunteering for the general population and for migrants specifically. We discuss research with respect to the two identified goals of volunteering as a policy tool, employability and empowerment.
The conclusions drawn from research on the effects of workfare volunteering on empowerment are very contradictory. Some argue that workfare volunteering disempowers people by depicting them as passive citizens, making them feel dependent and controlled (Andersen 2002), at the expense of their own sense of agency (Bessant 2000;Owen 1996) and autonomy (Warburton and Smith 2003). This is mainly due to the requirement (Levy 2006), but also to the lack of freedom of choice (Warburton and Smith 2003) and voice (Knijn and van Berkel 2003). Other critical studies conclude that workfare volunteering can be disempowering, since it is disciplining (Cruikshank 1999;Dean 1995Dean , 2010Marston 2005;Rose 1999;Walters 2000), restricting individual freedom by weakening agency and narrowing the ideal of citizenship to that of the worker-citizen (McDonald and Marston 2005;Warburton and Smith 2003). These authors often conclude that citizens become instruments of governance, instrumentalized and exploited to do unpaid work and to even feel happy about it: 'Power creeps under the skin of those who are governed to such a degree that they are even happy to conform' (Cruikshank 1999, p. 10).
In contrast, other researchers have found that volunteering makes welfare clients perceive themselves as active citizens, which bolsters their confidence and sense of control over their life (Cohen 2009;Koen 2013), with recognition from outside the labor market alleviating social marginality (Fuller et al. 2008).
There has been less research about the effect of volunteering on the empowerment of migrants specifically. An informant in the study by Handy and Greenspan indicated that volunteering was a way to show that he was a responsible and good person (2009: pp. 972-973). Moreover, respondents mentioned the opportunity to create a better image of their ethno-religious congregation in the wider community by being active as volunteers (Handy and Greenspan 2009). This last point relates to the findings of Fuller et al. indicating the potential role of volunteering in alleviating social marginality (2008).
Research on the effect of workfare volunteering on welfare clients' employability has so far hardly considered the consequences for citizenship. However, with regard to the effect of workfare volunteering on employability, the results are quite unanimous. Most research concludes that volunteering increases the employability of welfare recipients, both directly and indirectly. Volunteering is found to be a stepping-stone toward paid employment as it provides welfare clients with skills to find work (Cameron 1997;Cress et al. 1997;Reitsma-Street et al. 2000), and the opportunity to gain access to new networks and accumulate social capital (E. Cox 1997;Crick 2000;Giddens 2013;Pfau-Effinger and Herregaard 2006;Putnam 2000;Roberts and Devine 2004;Soupourmas and Ironmonger 2002;Turner 2001). Social capital here is defined as the social contacts and networks that provide opportunities and resources (Putnam 2000). Scholars usually distinguish between bonding and bridging social capital. Bonding social capital refers to the resources individuals can access through within-group ties, and bridging social capital refers to the resources individuals can access through ties that perform a so-called bridging function to outgroup resources (Adler and Kwon 2002;Kavanaugh et al. 2005;Leonard 2004;Putnam 2000).
Distinguishing bonding and bridging social capital is particularly important when considering migrant integration. Lancee found that cross-ethnic bridging social capital-and not bonding social capital-has a positive effect on employment and income among migrants living in the Netherlands (Lancee 2010). However, research shows that volunteering often leads to bonding social capital among migrants, but not so much to bridging social capital. A study on the volunteering of migrants in religious congregations found that volunteering contributes to strengthening social ties among their own ethnic-religious groups, but does not contribute to developing cross-ethnic social ties (Handy and Greenspan 2009). In a study about the role of volunteering in language attainment of migrants, Dudley (2007) found that while there were potential benefits of volunteering, most migrants only interacted with co-ethnics and other migrants, which hindered them to learn the language. In both cases, migrants mostly volunteer with co-ethnics or other migrants, and as such do not develop bridging social ties to the majority population which are important for finding paid employment (Lancee 2010). This would mean that volunteering is rather an obstacle than a route toward attaining Dutch citizenship ideals.
Other studies have found potential negative effects of volunteering on employability. Kampen (2014) has shown how the volunteer route, which is supposed to lead to paid employment, often ends in being stuck in an unpaid volunteering position with little prospect of paid employment. As a result, workfare volunteers turn away from the labor market and attach to their volunteer job (Kampen et al. 2013). Another study found that welfare recipients themselves often do not believe volunteering constitutes 'real' work experience, and doubt it will lead to paid work (Sawer 2006).
Regarding employability of migrants, most studies provide support for a positive effect of volunteering. Handy et al. (2009) found that volunteering contributes to developing different skills, gaining experience in the country of residence and in some cases functioned as 'a foot in the door,' which for some migrants resulted in getting a paid job (Handy and Greenspan 2009). However, the benefits for economic integration may vary. Recent migrants perceived more employment-related benefits of volunteering than established migrants (Handy and Greenspan 2009). Volunteering also contributes to language attainment of migrants. The extent to which volunteering contributes to language attainment depends on the opportunity the migrant gets to practice the language in a specific volunteering activity and the ethnicity and language skills of the individuals they interact with (Dudley 2007).
To sum up: the Dutch government uses workfare volunteering as a policy tool to stimulate migrant integration by enhancing their employability and empowerment. Earlier research on the ways volunteering contributes to or hinders employability and empowerment has shown us that workfare volunteering might affect migrant integration in numerous different ways.

Methods
In this article, we aim to answer the research question 'how and under which conditions does volunteering contribute to or hinder the employability and empowerment of migrant women living in the Netherlands?'. We held life story interviews (Atkinson 1998(Atkinson , 2012 because they are particularly suited to examine how migrant women experience the role of volunteering in the past, present and future, and to better understand the mechanisms determining the outcomes of volunteering throughout the life course. We focus on women of Turkish, Moroccan and Surinamese descent as these groups have disproportionately high unemployment rates compared to native Dutch and migrant men in the Netherlands (Huijnk et al. 2014) and are therefore considered to be most in need of empowerment and increasing their employability. We consider both the first-and second-generation migrant women because even though the second generation is slowly closing the gap, second-generation migrant women are still falling behind with respect to educational attainment and employment rates (Huijnk et al. 2010). Previous research found that recent migrants benefit more from volunteering than established migrants (Handy and Greenspan 2009). Including both the first-and second-generation migrants in this study, therefore, allows us to explore whether the first-generation migrants benefit more from volunteering than the second-generation migrants in the Netherlands.
Individuals of Moroccan, Turkish and Surinamese descent are the largest ethnic minorities in the Netherlands. 2 Most Turkish and Moroccan women arrived through family reunification joining their husbands who arrived as so-called 'guest-workers' and through family formation, marrying someone already living in the Netherlands (Entzinger 2006). Suriname is a former colony of the Netherlands and most women from Suriname arrived in the Netherlands around the independence of Suriname or to study (Entzinger 2006). Surinamese women were already more familiar with Dutch culture and the Dutch language before arrival, are predominantly Christian and in general have a better socioeconomic position in the Netherlands (Carabain and Bekkers 2011). Turkish and Moroccan women, on the other hand, did not speak the language at arrival, are predominantly Muslim and are from lower socioeconomic background compared to women of Surinamese origin (Carabain and Bekkers 2011). The women differ in their cultural and migration background, yet share their weak position on the Dutch labor market compared to native Dutch and migrant men (Huijnk et al. 2014).
Throughout 2015, the first author held semi-structured life story interviews (N = 46) with women of Moroccan (N = 19), Turkish (N = 19) and Surinamese (N = 8) descent. We used purposive sampling to select women living in the four largest cities of the Netherlands. 3 We recruited respondents through various methods. Firstly, flyers and posters were spread among various associations, community centers and Dutch language schools. Secondly, we held short speeches to introduce the research project, answer questions about the project and recruit respondents. 4 Thirdly, we used a snowball method to find women in communities that were particularly difficult to reach. The data were collected by the first author as part of a broader research project about the health and participation of ethnic minority women. Due to the wider focus of this research project, only respondents were selected who reported to sometimes have headaches and/or shoulder/ back/neck pain. As these are the most common health complaints among women of Turkish, Moroccan and Surinamese background living in the Netherlands (Hessing-Wagner 2006;Lindert et al. 2004), this selection criterion is unlikely to influence the generalizability of the findings. In the selection of respondents, we aimed to optimize the representativeness of the sample with respect to ethnic background, age, migrant generation and educational background.
The respondents are between 26 and 55 years old, with an average age of 39 years old. From all respondents, 48% were married, 16% divorced, 6% widowed, 14% has a partner and 16% is single. Among Moroccan, Turkish and Surinamese women, respectively, 47, 68 and 38% were from the first migrant generation. The majority of the women were currently unemployed, only 26% of the women currently had a job. 48% of the women completed up to secondary education (see Table 1 in the supplement for more detailed information). All women we interviewed have at some point been active as a volunteer. 30% of the women were currently not volunteering anymore, we still included these women in the analysis as their experienced motivations and outcomes of volunteering contribute to answering our research question. 52% of the women indicated the (indirect) involvement of government officials in starting volunteering. Volunteering was either a part of an integration or language course they were taking, a part of an activation program or workfare volunteering.
In the life story interviews, women were asked to, among others, discuss their motivations for experiences with and outcomes of volunteering and labor market participation. Three women made use of an interpreter (at their request). The interviews lasted from about 1 to 3 h, which were audio-recorded, transcribed ad verbatim, coded using Atlas.ti and were conducted in accordance with ethical standards. 5 We analyzed the interviews in Dutch with a specific focus on how firstand second-generation migrant women experience the effect of volunteering on employability and empowerment. Quotes were translated into English after selection in close cooperation with all co-authors. 4 Two of the organizations had a specific ethnic minority focus; Tori Oso focuses on Surinamese individuals and ATKB focuses on Turkish women living in Amsterdam. 5 The data were collected in accordance with the ICC/ESOMAR International Code on Market and Social Research and the Code for the Use of Personal Data in Scientific Research as formulated by the Association of Universities the Netherlands. Elaborate measures were taken to ensure the privacy of the respondents, and informed consent was obtained from all respondents. According to Dutch legislation, researchers are legally not obliged to request ethics approval (Dutch Civil Law, Article 7:458). In conclusion, ethics approval was not required for this study, yet this study lives up to a high standard of ethical requirements.

Results
Workfare volunteering policies are aimed at contributing to empowerment and employability. Below we discuss how these goals are achieved in the eyes of the migrant women we interviewed and which type of volunteering experiences are particularly conducive of contributing to these goals. We present the results in two major sections, empowerment and employability. We pay specific attention to the role of social capital in both sections since we found social capital to be an important determinant of both empowerment and employability of migrant women.

Empowering Experiences
Workfare volunteering is intended to contribute to the empowerment of migrant women. The interviews with the women show how empowerment is in fact an important outcome of (workfare) volunteering. We will present empirical examples from the stories of Arhimou, Rowena and Aliye, three examples of women representing three patterns of empowerment as an outcome of workfare volunteering. However, the example of Aliye also shows how easily empowerment is threatened. The example of Naomi, which we will present at the end of this paragraph, shows when workfare volunteering results in feelings of disempowerment. The empirical examples offer us five conditions that need to be present for volunteer work to empower migrant women.
The first condition under which volunteer work empowers migrant women is the possibility of putting former education or experience into practice. Arhimou, a Moroccan woman of 46 years old, has always been a housewife but recently started working as a volunteer in a sowing group. She completed a course to become a seamstress before she got married back in Morocco but after her marriage she did not have the opportunity to use these skills. Before the following excerpt she proudly tells how she is able to help other women in the sowing class. She then goes on to explain how her confidence grew because of her volunteer work: Luckily I am able to do all these things, as a little handy man haha. Yes really, when I am there, I think sometimes, it is good that at least I took this sowing course when I was young. Then at least I have really accomplished something for myself. Yes, it is really a lot of fun. It is also mine, it is.. I always dream about really doing something like that, with sowing […] Then I think yeah, some women go to work, everyone does something, but then I used to think, I am just sitting at home all by myself. That I really don't like. But then if I look at my life right now, then I'm really happy.
The opportunity to use her skills makes Arhimou feel appreciated and valued. This example shows how volunteering in itself can be a beneficial experience.
It is worth noting that in order to empower volunteer work does not necessarily have to improve someone's employability. In Arhimou's case, volunteering does not contribute much to language attainment (the women in this group also speak predominantly Moroccan), learning new skills (she already knows how to sow) or finding paid employment, it does provide Arhimou with a source of meaning and purpose, and the feeling of being valued.
A second condition under which volunteer work empowers migrant women is when it allows them to positively reframe their self-image. An example of a woman for whom volunteering is such an empowering experience is Rowena. Like Arhimou, Rowena is also able to use her former education, but more importantly, volunteering allows her to transform what she believes are her 'weaknesses' into 'strengths.' Rowena is 28 years old and of Surinamese descent (second generation). She is currently studying at a university, but missed many years of study due to a severe car accident at the age of 22, which resulted in a lengthy coma, brain damage and chronical physical pain. She tells about how she experiences volunteering at a foundation for people with similar conditions. My condition always stood in the way of my studies, like water and fire. But here, when volunteering, they are both assets, you see? It is difficult to combine, but it also gives me energy. At the university I sometimes feel like the difficult one, the one who needs more time, who needs special help. I always get it, but I don't want to need it. Before I never needed it, so.. Well, here they know about my experience and here it's an asset, that's why I got the job, they value it because I have first-hand experience with it [having brain damage].
Volunteering allows her to use her major obstacle in life, her condition as a result of the car accident, as an asset. This particular volunteer job allows her to transform herself from 'the difficult one' into 'the expert by experience.' This has an empowering effect, since it gives her strength and contributes to her confidence and self-esteem.
The third condition under which volunteer work empowers migrant women is when it provides them the peace that enables them to work at their own pace and allows them to make mistakes. This resonates with previous research that found that volunteer work offers welfare clients an empowering 'respite' (Kampen 2014). Aliye, a Turkish second-generation migrant woman who suffered from depression after losing her husband and lost her job due to budget cuts, describes clearly how the absence of pressure enables her to build up confidence and self-esteem.
Well, I wasn't really ready for a job, because a paid job also comes with extra pressure […] You have to achieve. So yeah, that's why I don't mind so much if I make a little mistake here. But when I just started here I was very insecure. I was always like oh no, am I doing everything right? But they are very relaxed here. They said yes, you are doing everything really well. Don't worry just do your thing. Yeah, so yeah, that really helped me.
Aliye highlights how the absence of pressure at her volunteer job contributes to feeling empowered.

Empowering Social Capital
A fourth condition under which volunteer work empowers is when it contributes to migrant women's bonding social capital. Most of the organizations where our respondents volunteer are based in specific neighborhoods, often the less well-off neighborhoods in the bigger cities. As such, the people our respondents meet often have similar educational backgrounds and financial situations. Latifa describes how the people that work the community center where she volunteers help each other: Many people around here struggle. They have no job, depend on social welfare, or have financial issues. But that is also the good thing, we all understand that. If someone can't make it because there is no money left to pay for a bus ticket to the community center, then someone else picks her up, like that. We've all been there, tough times, so it is good we can support each other here.
Latifa's remarks highlight a general pattern. Samira, a Moroccan woman who helps out in a sowing group, explains that language is a binding factor, because it facilitates understanding: Yes all of them [are Moroccan], some don't speak Dutch, so it is better, we all understand each other. The Monday afternoon group is more mixed, there they speak Dutch, so it is easier like this, we can all understand each other.
Hence, women become segregated into groups with members from a similar background.
In the two cases described above, volunteering provides limited access to bridging social capital. This homophily, the preference for individuals similar to oneself, is not surprising and is a well-known phenonemon in sociological research. The volunteer scene is no exception to the rule that individuals with similar backgrounds and similar interests tend to flock together (McPherson et al. 2001). Despite that these types of social ties provide little additional access to resources outside of the in-group, the migrant women highly value having bonding social capital. Understanding each other, being able to be yourself and providing support in case of difficulties are benefits several women emphasize. However, this way volunteering may provide bonding social capital, but not the bridging social capital which has been found to be associated with finding employment (Lancee 2010). Without an increasing chance to return to the labor market, volunteer eventually disempowers.

Disempowering Expectations
The previous examples show how and under which conditions volunteering leads to empowerment. However, about a third of the women have the opposite experience. When the abovementioned conditions-the possibilities of using skills, reframing self-image, absence of pressure, or bonding social capital-are no longer present, volunteering may result in disempowerment. This happens mostly when volunteering does not contribute to the prospect of having a paid job. In several cases we noticed how empowerment was contradicted by increasing expectations, such as the expectation of finding paid employment. To Aliye, for instance, volunteer work and paid work are not the same thing. Quite the opposite: she feels volunteer work offers her an escape from the pressure of doing paid work. However, her caseworker at the welfare office seems to think otherwise and puts pressure on her.
But now welfare officials say 'oh you are volunteering, then you can also have a paid job'.
[…] But this is really my own thing. I think it is really weird if you punish people who are actively trying to get back into that work rhythm, who want to do something valuable.. Yeah, also to become a bit more secure of themselves.
The pressure welfare officials put on Aliye to find paid employment contradicts the empowering effect of the absence of pressure at her volunteer job. The focus on paid employment seems to overshadow her accomplishment of volunteering as a way to deal with losing her spouse and suffering from depression, instead of encouragement she feels like she is being punished for her efforts. The pressure by Aliye's caseworker disempowers her.
A second reason why volunteer work disempowers migrant women is because not being paid may eventually contribute to feeling used. So in the long run there is a fifth condition that needs to be present in order for volunteer work to empower migrant women: eventually it needs to pay off. Naomi, a Surinamese firstgeneration migrant, has recently lost her job and picked up volunteering with the vague promise of paid employment at the organization where she volunteers. At first, helping people empowered her, but eventually the feeling of being used gets the upper hand: Yeah I really don't mind [working unpaid] for a few months, but if they think you are so great and you have been working there for a while and then they say 'oh sorry we can not pay you' but they do want you to do all kinds of things and work really hard… (…) Then I feel a bit like, that they are actually using you, yeah, hahaha! I mean, it is great to help people, but you can only help others if you can sustain yourself.
Especially the idea of being taken advantage of seems to disempower Naomi. She explains why it matters who is profiting of her kindness: You know, I really don't mind volunteering for people who have nothing, the poor, the elderly, the disabled, but for organizations who have tons of money but they just feel like not paying you.. that I really don't like. Then I really don't feel valued you know. You have to do volunteer work from the heart. And not something like, ok, we think you do a great job but we are not going to pay you.
For Naomi, the experience of volunteering for an organization with the means to pay her is disempowering. She feels valued when helping others, but only under the condition of being in a position of financial security.

Employability
An important goal of workfare volunteering is increasing employability. Employability should of course eventually lead to finding paid employment. Indeed, a recurring theme in many of the interviews is how volunteering might increase employability by allowing the women to practice Dutch, apply their skills or learn new skills. In general, for individuals with elaborate work experience and sufficient educational attainment, volunteering seems to have less of an added value for developing employability than for individuals who lack skills and experience. Volunteering should therefore be mostly beneficial to migrant women with a large 'distance to the labor market.' However, Dutch employers strongly focus on diplomas and certificates, which results in our informants' newly gained skills not being formally recognized and then ultimately not getting a job. The lack of formal recognition of their skills and experience may even result in feelings of disillusionment and disappointment. This, we will argue, ultimately leads the migrant women to feel disempowered. We will discuss four conditions that need to be present for volunteer work to improve the employability of migrant women.
Volunteering contributes to migrant women's employability, but only under certain conditions. The first and most fundamental condition under which volunteer work contributes to the employability of migrant women is being able to interact with people at the volunteer job. Many of the women we talked to mention improving their Dutch as one of the reasons for volunteering and that being able to practice while volunteering contributes to improving language skills. An example is Gülay, a Turkish woman who volunteers at an elderly home. She explains she started volunteering in the elderly home to practice Dutch. ''I wanted to talk, it became better, I wanted it myself. I wanted to practice my Dutch, and then also my Dutch improved.'' However, the example of Irmak, a recently migrated Turkish woman, shows how language attainment depends on being able to interact with others at the volunteer job.
I did a 'language volunteering internship' at Zeeman [a budget clothing store]. But I don't like it. I fold clothes, all day, folding, folding, much work. No one talks to me. They say 'you go fold', 'you go clean'. How practice Dutch when only folding clothes?
As a result, migrant women that have been volunteering without practicing the Dutch language do not believe they have more chances of finding a paid job. For them, the volunteer organization remains their main focal point for future employment, since they have at least proven their value there. Gülay responds to the question if she would like to work: ''Work? Me? Ehh, look, my Dutch is a problem. Still. But, I do want to. (…) Maybe in the elderly home, like what I do as volunteer, helping old people. (…) But I don't know how.'' The second condition under which volunteer work contributes to employability is a challenging volunteer activity. The example of Khadija, a Moroccan mother of five who has always been a homemaker, shows how skills attainment depends on the familiarity with what migrant women are expected to do. Being too familiar results in the volunteer work being too less of a challenge.
My participation coach said I should do something, get out of the house. He told me about this women's center, so now I am here. I do the dishes, clean up, clean the toilets.. Yes. Well, I am out of the house but I do the same things I do at home. I see some of my neighbors here, then we drink tea. But I don't really understand why he wants me to go here.
Khadija's example shows that having challenging volunteer work can be decisive for acquiring skills. Khadija more or less does the same things she has always done at home. The activities are not challenging enough and thus do not contribute to Khadija's employability. On top of that, she does not acquire new social capital, since she meets her own neighbors at the women's center. However, the following examples make clear that even when migrant women acquire bridging social capital, this does not guarantee employability.

Bridging Toward Employment
The third condition under which volunteer work leads to increased employability is the opportunity to build bridging social capital. Nabila, a first-generation marriage migrant from Morocco, tells us about the importance of her volunteer work for practicing the language and learning new skills. Her story is exceptional with respect to her access to bridging social capital. Nabila received a lot of support with finding a job, but in the end she did not manage to find a paid job. At her volunteer job, she met Silvia who taught her about administration and organization skills: ''She really took time to explain how things work, e-mail, making the agenda on the computer, taking minutes at a meeting. If I made a mistake she just helped me.'' Nabila appreciates the help, but she explains to Silvia she has to move on, because she needs to earn money. Silvia helps her: She found all these jobs online and send me this website. She also helped me write a letter and to write down my experience from the volunteering. She even put her name down as a reference. Really really nice of her.
The conditions at Nabila's volunteer job seem optimal in many respects: she's got access to bridging social capital, she receives support and coaching, she is taught many new useful skills, gains experience and receives practical support in finding and applying for a job. However, even this optimal volunteering experience does not result in a paid job.

Still Unemployable
Continuing with the example of Nabila, even when volunteering provides opportunities to build bridging social capital and to learn the language and useful skills, still the route toward paid employment is not secured when this experience is not recognized as a qualification for the labor market. Even with Silvia's help Nabila is rejected time and again: ''So, I start applying, applying, applying, so many letters I have send. But they never want me. I don't even get a response.'' The reason for being rejected is particularly painful to her, because it misrecognizes her experience as a volunteer: When I call they say: 'Sorry there is someone with a diploma, with 'real' work experience'. Real? Like what I do is not real? Do I make fake phone calls? Or fake administration? Well, I got fed up with those rejections. I just kind of stayed on at the school, but different than before, I just help a bit.
While not having the proper diploma's Nabila is rejected repeatedly, but the fact that her volunteering experience is not perceived as 'real' work experience is most painful to her. The only way to be recognized as having 'real work experience' is by getting a diploma. Frankly, volunteer work does not lead to getting a formal diploma. If a diploma is a prerequisite to enter the labor market, volunteer work can never be the stepping-stone it was promised to be.
Despite her raised self-confidence, acquired skills, improved language and bridging social capital, employers still consider Nabila 'unemployable.' Ironically, this results in disempowerment, since despite all her efforts and accomplishments she ends up 'fed up with rejections' and eventually even downplays her efforts as a volunteer as 'just helping a bit.' So, it seems that the fourth and most important condition under which volunteer work leads to increased employability is a formal diploma. This condition even proved to be important when already hired, because having a diploma proved to be an important condition to remain employed. The only informant who found a paid job through volunteering is Elbahia. Elbahia is a Moroccan woman who migrated to the Netherlands as a child. Regarding her employment history she told us: ''I was working first as a volunteer at the elderly home, but I was doing very well so after 1 or 2 years they offered me a paid job.'' However, the need for diplomas at her employer became a source of frustration and even resulted in Elbahia quitting her job.
When going on pregnancy leave from her paid job, after working there for more than 15 years, her former intern gets promoted to become Elbahia's supervisor. She explains: Look, she had a diploma, and I didn't. I did the same job without a diploma. But I did it really well, everyone says so! My colleagues, the clients.. there were even clients that only wanted to be helped by me! But my intern did have that piece of paper, so suddenly she was placed above me.
This made Elbahia angry and motivated at the same time: So I suddenly had to obey my intern?! And that is not even what I was angry about, it was how secretive they were about it, they didn't even tell me. Anyways, when they finally told me I was like: 'That is all fine, but apparently it's all about a piece of paper? Then I also want to get that little piece of paper!' When Elbahia graduates she is still so hurt by the way she was treated that she decides to quit: Voluntas (2017Voluntas ( ) 28:1900Voluntas ( -1921Voluntas ( 1915  This example shows how at the same time recognition and a degree are important conditions to stay employed, even after working at the same employer for over a decade. The preference that is given to someone with less experience but with 'that little piece of paper' felt as being misrecognized for her efforts. The humiliation has an empowering impact on Elbahia, because it motivates her to prove her employer wrong. However, this type of empowerment does not add up to her employability for the particular job she was doing. On the contrary, she feels she is treated so unfairly that she decides to quit after proving to be able to formally qualify for the job. Ironically, Elbahia ends up more employable in a formal sense, but without employment.

Discussion
Workfare volunteering is a Dutch policy measure aimed at promoting migrant integration. So far the effect of volunteering on migrant integration has been underresearched. Drawing on citizenship theory (Newman and Clarke 2009), we studied whether, and under which conditions, workfare volunteering contributes to the integration of migrant women by impacting their empowerment and employability. Drawing on citizenship theory, we have determined that the policy is based on two images of the ideal citizen: 'the empowered citizen' and 'the worker-citizen' (Newman and Clarke 2009).
Volunteering particularly contributes to the empowerment of our informants when they were able to use their skills or reframe their self-image. Also, finding purpose and meaning in volunteer work raised their confidence and self-worth. Volunteering empowered some women by enabling them to envision themselves in a paid job. However, at the same time the ideal of the 'worker-citizen' disempowered these women, since it turned out to be an unattainable ideal. Our respondents repeatedly told us how employers do not consider their volunteering to be real work experience and do not recognize their new skills and knowledge when not formalized by a diploma. As such, volunteering only resulted in finding paid employment for one of our 46 respondents. At the same time, welfare officials idealized the worker-citizen image by pressuring our respondents to find paid employment. Not being able to live up to this ideal resulted in disempowerment. This was worsened when the women felt like being used as 'free labor.' Ironically, disempowerment made some women give up looking for a job or quit their volunteer work altogether.
Our novel approach shows how striving toward different policy goals at once affects the outcomes of workfare volunteering. The main reason for the negative outcomes of workfare volunteering is the two citizenship ideals of the empowered and the employable citizen. These citizenship goals contradict each other because these goals require different, often contrary, conditions in order to be attainable. These findings empirically complement Newman and Clarke's (2009) observation that several images of ideal citizens 'swirl around' and show how this may result in citizens ending up empty-handed. In the case of workfare volunteering by migrant women, increasing empowerment and employability at once turns out to be counterproductive.
An increasingly demanding definition of migrant integration (Vasta 2007) in combination with the process of mainstreaming (Entzinger 2014) has resulted in rising expectations of migrants. Our findings show that raising integration expectations are not only counterproductive but may also eliminate previous gains. Migrant women are empowered up to the point that they have proven themselves to be fit to enter the labor market and at that point they are told their work experience is not real enough to participate as a true Dutch citizen. This means that the Dutch citizen ideal of the worker-citizen is never actually feasible for them while being on welfare, which also makes it harder to live up to the ideal of the empowered citizen.
There are some limitations to our study. Firstly, we interviewed the informants at one point in time. The way they narrated their motivations for, experiences with and outcomes of volunteering in the past may therefore be influenced by their current situation. Yet, simultaneously, these retrospective accounts also provide richer insights. Most respondents narrated both how they understood their motivations at a certain time point but also how their understanding of their motivations progressed and developed over time. Making statements such as 'At the time I thought I started volunteering to get a job, but actually, when looking back, I was just lonely'. Still, we believe that the research on the effect of volunteering on integration will benefit from longitudinal designs.
Secondly, due to the qualitative nature of the data we cannot claim causality. However, how individuals subjectively perceive the impact of volunteering on integration provides valuable information. Moreover, we found that most women responded negative to the question whether they were active volunteers. Yet, when asking about their daily activities, it appeared most were actually volunteering. The term 'volunteering' seems to be scarcely used in these groups, women called it 'helping out' or used a more specific description to describe their roles. Survey questions about volunteering would have resulted in bias in the results. Future studies should use carefully designed questions directed at measuring volunteering due to a limited use of the term 'volunteering' among certain groups.
Thirdly, the specific nature of the sample may limit the generalizability of our findings. Day and Devlin, for example, found that volunteering resulted in an 11% increase in wages for men, while volunteering had no positive effect on wages for women living in Canada (Day and Devlin 1997). As we only included women in our sample, it is unclear whether our findings are generalizable to migrant men. Moreover, we focused on women of Turkish, Moroccan and Surinamese descent living in the Netherlands. Are our findings also generalizable to other ethnic groups and different national contexts? Firstly, our findings confirm previous findings conducted across various ethnic groups in a different national context. Handy and Greenspan found that volunteering in ethnic congregations contributed to developing social capital and human capital (employability) in a study among migrants from various ethnic backgrounds, though predominantly Asian, in four Canadian cities (2009). Dudley found that volunteering contributes to language attainment and social integration in a sample of migrants coming from 24 different countries living in Canada (2007). Our studies indicate that despite the diversity in ethnic backgrounds of the migrants and the different national contexts, similar beneficial effects of volunteering on migrant integration are found. This strengthens our confidence in the generalizability of our main findings. However, due to the effect of individual and contextual characteristics on the outcomes of volunteering, we suggest more future research is needed to study under which conditions for which migrant groups volunteering may have positive effects on integration.
Despite the various positive outcomes of volunteering that we found, our results show that workfare volunteering may not be an adequate solution to unemployment and migrant integration. In order to make sure workfare volunteers benefit from the positive outcomes of volunteering, we will suggest some policy changes.
First, there is a role for organizations that offer volunteer positions to the unemployed. These organizations should formally recognize gained skills and knowledge by offering certificates that enable workfare volunteers to underline the value of their experience. Also, public sector organizations should strive for an ethnic and socioeconomic mix of their volunteers, to improve the conditions to develop bridging social capital. Private sector organizations that benefit from the free labor of volunteers should pay them back by investing in volunteers. For instance, Private sector organizations could offer on-the-job training in relevant skills and could officially recognize experience by issuing certificates. After a period of time, private sector organizations that employ volunteers should be held to offer them a paid position.
Second, welfare officials have an important role in the process. They should be aware of the many obstacles on the path to paid employment. Often preferences among employers block the individual's route to paid employment. Welfare officials should focus on in-between victories instead of solely on paid employment in order to paint a more realistic picture of welfare clients chances on the labor market.