Unpacking intersectional solidarity: dimensions of power in coalitions

Scholars have often oscillated between celebrating the transformative potential of solidarity and recognising the ambivalent nature of cooperation for disadvantaged and marginalised groups. How can we make sense of these differences? This article addresses this question by unpacking intersectional solidarity along two dimensions: the ways issues are framed; and the extent to which organisations adopt a transformative praxis to redress disparities in resources and representation. By focusing on the interplay between discursive and material dimensions of power, we identify four types of intersectional solidarity, with different transformative potential. The usefulness of this typology is illustrated by means of secondary analysis of coalition work developed around reproductive justice and domestic workers’ rights. It shows that only an ideal form of transformative solidarity reflects feminist normative theorising of an alliance across differences.


Introduction
Feminism has a long tradition of building networks and alliances across national boundaries and racial, ethnic, class and sexual divisions. These alliances have often proven effective in heightening the political influence of feminist actors and promoting inclusive transformative projects, but they have also often led to the marginalisation of minority issues and the co-optation of less powerful organisations (Lorde, 1984;hooks, 1986;Alvarez, 1999;Batliwala, 2002;Mohanty, 2003;Roth, 2004, Strolovitch, 2008. There is an ambivalence in the concept of solidarity that stems from lumping together phenomena with very different characteristics, fields of applicability, causes and outcomes. The aim of this article is to offer a conceptual framework of intersectional solidarity that puts at the centre of the analysis the relationship between diverse social struggles and inequalities. Definitions of solidarity have varied greatly across fields of the sociopolitical sciences. Here, we focus on solidarity as an ongoing political process of creating ties and coalitions across social differences by negotiating power asymmetries (Tormos, 2017). Social movement scholarship views solidarity through the lens of coalitions (Staggenborg, 1986;Whittier, 2014;Van Dyke and Amos, 2017). This scholarship offers one of the most systematic theorisations of the process of coalition building, but intersectionality has not figured prominently in these studies. By theorising solidarity as coalition, it also ends up promoting a strategic and instrumental view of solidarity, which obscures more diffuse as well as voluntaristic and affective forms. Solidarity is also at the centre of recent studies of intersectional movements' praxis (Luna, 2010;Tormos, 2017;Ayoub, 2018;Price, 2018;Einwohner et al, 2019;Evans and Lépinard, 2019;Irvine et al, 2019a). This literature has greatly advanced our knowledge about the practices used by social justice movements to forge alliances but tends to treat solidarity as unidimensional, something that is either more or less present, active or passive, thus leaving untheorised the grey areas. While social movement scholars have identified a number of typologies of coalitions and the various political goals they serve, little attention has been given to this question in the literature on intersectionality (Irvine et al, 2019b). Drawing on these scholarships, this article proposes a framework for conceptualising distinct forms of intersectional solidarity along two dimensions: (1) how differences within and between organisations are reflected in the framing of issues; and (2) the extent to which organisations actively seek to redress power disparities in resources and representation. This framework has several advantages. First, it makes visible the need to consider the interplay of discursive and material dimensions of power in the analysis of intersectional solidarity. Second, it identifies four distinct types of intersectional solidarity, with different transformative potential. Third, it advances an ideal form of transformative solidarity as a way to contribute to normative theorising about the good life (Lyshaug, 2006;Rai, 2018).
The article is organised as follows. It begins by describing a number of challenges that confront the theorisation of intersectional solidarity; it then proceeds to draw out the key dimensions of solidarity and presents a typology. In the next section, this typology is put to use by using secondary sources to illustrate different types of solidarity and how they developed around issues of reproductive justice and domestic workers' rights in interactions between mainstream feminist organisations, organisations of women of colour and other collective actors (LGBTQ movement, migrant organisations and trade unions). We selected those issues because they often represented contested sites within feminism, with conflicts pointing to the demarcation between gender and other social divisions of class, race, sexual orientation and gender identity. Our purpose here is foremost one of conceptual development and clarification, but this exercise is also helpful to explore how particular combinations of discourses and practices shape intersectional solidarity.

Challenges in conceptualising feminist intersectional solidarity
The concept of intersectionality originates in the critique of black feminist activists of the tendency of mainstream feminist organisations to assume a unitary notion of women that suppresses identities and issues that lie at the intersection of gender and race (Combahee River Collective, 1977;hooks, 1986;Crenshaw, 1991;Collins, 2015). Since then, intersectionality has travelled a long way, becoming a veritable buzzword not only in the academic world, where we have witnessed a rapid growth of studies addressing diverse research questions and incorporating a wide array of inequalities (Davis, 2008;Mügge et al, 2018), 1 but also in the practices and discourses of social movement organisations (Evans and Lépinard, 2019;Irvine et al, 2019a). Political intersectionality, in particular, refers to two interconnected issues: ensuring that underprivileged groups are represented, empowered and given voice within organisations; and building coalitions across social divisions (hooks, 1986;Crenshaw, 1991). The first issue has received a great deal of attention; here, intersectionality is defined as a strategy, a repertoire of action and a set of practices used by feminist movements to counter power imbalances that lead to the marginalisation of disadvantaged groups within organisations (Weldon, 2006;Strolovitch, 2008;Chun et al, 2013;Lépinard, 2014;Einwohner et al, 2019). The second issue has generally received less attention, but recent years have witnesses a rapid increase in the number of publications looking at coalition politics using intersectionality as an analytical framework (Ayoub, 2018;Evans and Lépinard, 2019;Irvine et al, 2019a). We define this as intersectional solidarity, and it comprises instances of cooperation occurring between two or more organisations mobilising different constituencies defined by gender, sexuality, ethnicity/race, class and other divisions (Evans and Lépinard, 2019;Irvine et al, 2019a), or located across national boundaries, as in transnational cooperation (Weldon, 2006;Ahrens, 2019). To be sure, these two forms of viewing political intersectionality are not unrelated, and Erica Townsend-Bell's (2009; formulation of intersectional praxis probably represents one of the clearest formulations of the nexus between the two. As stated by Townsend-Bell (2009: 3), a commitment to intersectional praxis requires that: groups attend to all relevant axes of difference. Ideally, groups do this first within their own organizations, by engaging in a continual process of coalition that constantly reshapes the boundaries of the epistemological and ontological community(ies). They will then move on to create these kinds of alliances with other organizations, because their final commitment is to the eradication of power differences and marginalization.
While this definition describes an ideal of intersectional solidarity, several less advanced forms are also possible.
The study of feminist intersectional solidarity needs to confront two main challenges. The first deals with the existence of power asymmetries between organisations. Previous research has tended to oscillate between celebrating the transformative potential of solidarity and recognising the often-ambivalent nature of cooperation and the risk of exploitation of less powerful constituencies. The importance of considering power asymmetries has been particularly highlighted by black and Chicana theorists, as well as postcolonial and non-Western scholars (Anzaldúa and Moraga, 1981;Lorde, 1984;hooks, 1986;Grewal and Kaplan, 1994;Mohanty, 2003;Roth, 2004). Studies of transnational cooperation between organisations from the Global North and South, and between white and racialised groups, show that while powerful organisations are invited to partner with state actors and international organisations to lend their expertise, other groups often lack the institutional access and resources to enter these arenas, which leads to the silencing of critical voices (Alvarez, 1999;Batliwala, 2002). Such power asymmetries negatively affect intersectional solidarities, obstruct cooperation and distort agendas.
The second challenge in studying feminist intersectional solidarity involves the existence of inconsistencies and ambiguities around the definition of both feminism and solidarity. There is a long-standing debate on the essential features of feminist movements as distinguished from the more general concept of women's organisations, which also includes conservative women's groups. There is a certain consensus that a distinctive trait of feminist movements lies in the goal of transforming gender relations that subordinate women to men (Ferree and Mueller, 2004;Beckwith, 2007). However, narrow definitions focusing on single dimensions of inequality run the risk of privileging mainstream feminist organisations, concealing the contextual nature of feminist goals, as well as their intersection with other social justice struggles (Garcia, 1989;Ferree and Mueller, 2004;Montoya, 2014;Collins, 2015). Therefore, studies of intersectional solidarity should adopt a broad definition of feminist movements that also includes organisations mobilising around a larger set of issues (for example, racial justice and workers' rights), as long as their claims contain elements that clearly intersect with gender issues (Evans and Lépinard, 2019).
A further layer of complication in defining feminist movements relates to the fact that the boundaries of social movements are often vague. Social movements are networks of organisations that may support or form alliances among themselves (Zald and McCarthy, 1980;Della Porta and Rucht, 1995). While coalitions can sometimes be movement-wide, they are more often confined to specific organisations and activities. Feminist organisations are also internally heterogeneous. They are always composed of individuals who have multiple identities located at the intersection of interlocking forms of domination (Crenshaw, 1991) -what Carastathis (2013) calls 'intersectionalities within'. The way in which organisations deal with these differences -and the conflicts and antagonisms they generate -already opens or forecloses the potential for intersectionality solidarity.
The concept of solidarity is instead often used in vague and ambiguous ways. Several concepts are commonly adopted to refer to feminist solidarity, such as sisterhood, cooperation, alliances, coalitions, advocacy networks and feminist triangles (Woodward, 2003;Cole, 2008;Strolovitch, 2008;Hancock, 2011). While these concepts are sometimes used to refer to specific forms of collaboration, they are often used rather interchangeably. Furthermore, studies often emphasise single dimensions of solidarity in relation to the research problem at hand, but there is very little systematic analysis of how particular forms of solidarity differ from each other and of the conditions that facilitate or hinder their formation. Therefore, our analysis starts by investigating the concept of solidarity in order to identify key aspects underlying its different forms.

Theorising varieties of feminist intersectional solidarity
In this section, we aim to draw out key dimensions of the concept of solidarity in order to distinguish among different forms of intersectional solidarity. A promising starting point is the extant literature on coalitions. The literatures on both social movements and political intersectionality seem to agree that all movements are coalitions of sorts (Cole, 2008;Crenshaw, 1991;Carastathis, 2013;Van Dyke and McCammon, 2010). Indeed, the concept of coalition has a long history in both fields, where is often used to describe instances of collaboration between organisations. At the most basic level, coalitions are defined as instances where two or more social movement organisations work together towards a common goal (Van Dyke and McCammon, 2010). However, coalitions can take a variety of forms, which range from loosely coupled activities around some broadly defined issue, to the constitution of umbrella organisations that coordinate actions between members. Nancy Whittier's (2014; 2018) framework of movements' interaction offers a useful starting point to think about different forms of coalition. She distinguishes types of movements' relationship along two dimensions: (1) interaction or shared goal; and (2) congruence of ideology. The first dimension concerns how movements interact, which can range from working together to no direct interaction or opposition. In the analysis of intersectional solidarity, this dimension can be effectively omitted since, by definition, it excludes cases of no interaction and relationships of opposition. The second dimension points to what is generally considered in the social movement literature as an essential element of coalitions: ideological alignment (Van Dyke and McCammon, 2010). Here, ideological alignment is defined as a general compatibility of values and beliefs. Accordingly, radical and liberal feminists and labour and environmental movements are considered ideologically congruent because of their shared membership in a progressive social movement sector (Whittier, 2014). While we believe that this dimension captures some critical aspects of coalitions, this definition of ideology presents a number of problems. First, the definition of ideology in the social movement literature is far from clear and is used to refer to a number of different things, such as cultural congruence, similar collective identities or shared visions of social change and/or political ends (Van Dyke and Amos, 2017). The second, and more critical, limitation is that this use of ideology effectively considers the discursive alignment between organisations as a given rather than a product of coalition formation constructed in the back and forth between organisations.
Given these limitations of ideology, we propose to replace, or rather to narrow, its definition to the ways in which specific issues or goals are framed by political actors so as to deny, or rather acknowledge and counter, the existence of multiple inequalities. Despite the proliferation of scholarship on the importance of collective action frames in movements' emergence, strategies and outcomes, their role in coalition building has remained generally more implicit, and mostly limited to studies of feminist transnational activism (Weldon, 2006;Montoya, 2013). Nonetheless, cooperation between organisations requires an active process of negotiation and interpretation to build consensus around a particular framing of goals and issues. As stated by Ferree and Roth (1998: 628), it is through the framing of issues that movements and actors recognise, or rather distort and deny, the structural interconnection between inequalities in a system of oppression. Croteau and Hicks (2003) offer one of the few frameworks for the analysis of framing processes in coalition building, which they describe as the development of a 'consonant frame pyramid'. In the context of a coalition, framing is particularly challenging because organisations need to frame issues in ways that align with individual members' frames to ensure their support, as well as with the coalition frame to find some common ground for collaboration. This process is always imperfect and subject to disputes, and the more diverse the organisations involved, the more contentious and difficult the process. In this context, powerful organisations with more members, capacity and resources have greater ability to influence the coalition frame.
Based on the preceding discussion, our first dimension distinguishes forms of intersectional solidarity based on how frames generated and employed in cooperative interactions among organisations deal with differences within and between them. We identify two main ways in which goals/issues are framed: common denominator (CD) and recognition of differences (RD) frames.
CD frames occur when two or more organisations emphasise the commonalities of experience and interests between their members, and strive to identify solutions that will benefit all different groups within the coalition. Framing occurs through a process of identification of shared issues ('what we have in common'). Differences between members and organisations are addressed only to the extent that they serve to highlight cross-cutting issues relevant for all members of the coalition. Examples of the use of such frames are identifiable in the reliance on human rights discourses in the transnational movement against violence against women. The human rights frame defines violence as an issue (potentially) affecting all women and comprising a wide range of problems, from poverty, to female genital mutilation, to rape. The use of such a frame has been important in getting the issue on the agenda and building coalitions between North and South feminists and international human rights organisations (Keck and Sikkink, 1998;Weldon, 2006). 2 Nonetheless, this type of frame also conceals or underexposes specific forms of violence that are not cross-cutting, de facto silencing disadvantaged and minority groups (Schwenken, 2003;Strid et al, 2013). As CD frames emphasise universal issues that are shared by diverse inequality-based struggles, they are more vulnerable to being influenced by more powerful and resourceful organisations. Another potential disadvantage of CD frames is that their meanings can easily be diluted and/or co-opted (Hewitt, 2011).
RD frames focus instead on the incorporation rather than the minimisation of differences within and between organisations (Townsend-Bell, 2011). Hence, solidarity is built not around commonalities, but through a process of recognition of the specific nature that certain issues assume for minority and disadvantaged groups. Such frames are grounded in the knowledge produced by marginalised subjectivities and show an influence of minority and other less powerful groups on the framing process so as to highlight their unique challenges. RD frames are the product of a sort of 'dialogue across differences' in which organisations -especially dominant ones -begin to develop a critical awareness of their own biases and blind spots, and increasingly adopt the language, analytical categories and core concerns of other organisations in their frames. Luna's (2010) analysis of the 2004 March for Women's Lives offers an example of how RD frames may emerge. The author recounts how the participation of organisations of women of colour shifted the coalition frame from one of reproductive rights -woman's choice to have an abortion -to one of reproductive justice, explicitly acknowledging the different meaning that state control over reproduction has historically assumed for minority and poor people. Lucas Platero and Ortega-Arjonilla (2016) show, instead, that interactions between the Spanish trans and feminist movements have led to the emergence of a new intersectional 'transfeminist' frame focused on the criticism of the gender binary system. These works show that the production of RD frames may occur through the enlargement of a previously used concept (from reproductive rights to justice) or by making certain elements more prominent then they originally were (transfeminism). This type of coalition framing process requires that organisations act with an intersectional consciousness (Greenwood, 2008) and adopt what Erica Townsend-Bell (2011) calls an alternative 'politics of accountability', that is, a critical scrutiny of organisational practices and the use of inclusive forms of deliberation to counter the (re)production of inequalities both within and between collaborating organisations. We expect coalitions adopting RD frames to represent a more radical form of intersectional solidarity.
A second relevant dimension of intersectional solidarity comes from the analysis of the emerging literature on intersectional movement praxis (Weldon, 2006;Hancock, 2011;Montoya, 2013;Lépinard, 2014;Price, 2017;Einwohner et al, 2019;Evans and Lépinard, 2019;Irvine et al, 2019a). These studies have emphasised how a commitment to intersectional solidarity requires that organisations enact a repertoire of practices of inclusivity to ensure that less powerful groups are included on equal footing in shaping movements' goals, strategies and tactics. They range from making sure that there is a diverse leadership, to having inclusive mechanisms of deliberation and making changes in the ways organisations represent themselves. These practices aim to enhance the substantial representation of 'minority' issues by improving the descriptive representation of disadvantaged groups in the organisation. These studies point to an important element of intersectional solidarity, while also making clear the relationship between certain structures that aim to redress power dynamics between organisations (representation), and framing processes that recognise the specific claims of marginalised groups (recognition). However, drawing on Nancy Fraser's (1997) three approaches to remedying injustice, we believe that advanced forms of intersectional solidarity should also include an additional element: the redistribution of a wide range of resources, including financial, institutional and organisational ones. Advanced forms of intersectional solidarity necessitate a durable commitment to the abolition of oppressive relations beyond particular issues at hand (Townsend-Bell, 2009); they require actions aimed at the eradication of all forms of power asymmetries within a coalition so as to transform it into a collaboration among equals. We define this dimension as transformative praxis, which points to the need to eliminate disparities in both the ways in which coalition are organised and how resources are shared (Price, 2017). The inclusion of this second dimension emphasises how intersectional solidarity involves constant work by organisations to address power disparities at the symbolic level of how issues are framed, and an equally deep transformation of the material structures through which power works within coalitions. Table 1 categorises forms of intersectional feminist solidarity based on the two dimensions identified. Our preceding discussion made clear that they are not independent; rather, it is the way in which they interact and shape each other in particular contexts that gives rise to different forms of intersectional solidarity. Each dimension should be considered as a continuum, and their intersection identifies four ideal types: transformative, instrumental, pragmatic solidarity and incorporation. Differences between those forms vary based on the way that issues are framed so as to deny or rather acknowledge the problems affecting more disadvantaged groups, and the extent to which they adopt a transformative praxis to truly address power imbalances between groups. These types are rarely observable in their pure forms. For instance, organisations can resort to CD frames to move issues onto the agenda of policymakers but put greater emphasis on differences when it comes to framing solutions, as well as sometimes adopt different frames for external and internal audiences. 3 Thus, concrete instances of solidarity will often show elements of different types but their prevailing characteristics will still come closer to a specific type. In the next section, we look at instances of solidarity emerging around issues of reproductive justice and domestic workers' rights to illustrate the differences between those types and explore the conditions leading to more instrumental or rather advanced forms of solidarity.

Building intersectional solidarity around controversial issues
Based on the conceptual framework presented earlier, we use secondary literature to illustrate how intersectional solidarity (or the lack thereof) developed in interactions around issues of reproductive justice and domestic workers' rights. We selected those issues because they have historically been -and continue to be -divisive of feminist movements. Our interest here is on the interplay between discursive and material dimensions of solidarity, and how this leads to the different types of intersectional feminist solidarity identified in Table 1 given certain contextual conditions. Controversial issues offer excellent material to analyse these questions. Debates around abortion and sexual rights, as well as around the value of domestic and reproductive work as a path to women's emancipation, have always splintered the feminist movement (Kretschmer, 2014;Ciccia and Sainsbury, 2018). Not only is solidarity more difficult to manifest, and all the more needed, but the contested nature of those issues also implies the presence of less established frames and discursive disputes over the representation of the problem and desirable solutions. Moreover, as we will show, the demarcation between gender and other social divisions of class, race, sexual orientation and gender identity cuts across and underlines all these conflicts.
We begin by looking at instances of coalition work around reproductive justice to illustrate differences between instrumental and transformative solidarity (types 1 and 2). Reproductive justice is a movement -and a frame -initiated by women of colour that uses intersectionality, human rights and social justice ideas to broaden the mainstream reproductive rights framework and its singular focus on abortion (Ross, 2017). By emphasising the connection between reproduction and other social justice struggles, such as poverty, racial injustice, immigration, violence and prisoners' rights, it aims to broaden the scope of reproductive struggles to include the rights to have children and to parent those children (Price, 2010). The reproductive justice frame does not focus on identities, but rather aims to draw the connections between diverse social justice struggles related to reproduction in the context of complex systems of oppression (Ross, 2017). Zakiya Luna's (2010) analysis of the 2004 March for Women's Lives illustrates the limitations of instrumental forms of solidarity. The author documents how the framing of the march organised by four well-established national women's organisations shifted from one of reproductive rights to one of reproductive justice when Sister Song joined the coalition. Sister Song is a network of over 80 women of colour and allied organisations founded in 1997 that works to achieve reproductive justice. Various factors worked to promote the formation of a coalition, including the threat presented by the signing of the Partial-Birth Abortion Ban by President Bush and the increased criticism of mainstream feminist organisations for their lack of diversity. The construction of a coalition was not automatic and required negotiations between Sister Song and the four women's organisations on the name of the march over the content of documents that were issued, space on the planning committee and fee waivers worth US$250,000 each for membership of the steering committee. Despite the great success of the march in terms of numbers, the coalition was short-lived. While the mainstream organisations have maintained some reference to reproductive justice in their work, they have not integrated it into a deeper analysis of their organisation and continue to remain focused on abortion (Luna, 2010).
Coalition building took a more decisive transformative turn in the case involving the reproductive justice and LGBTQ movements documented by Kimala Price (2017;. The author recounts how the process of building a coalition spanned several years of meetings and conferences created explicitly with the intent to discuss the potential for collaboration. These encounters were not meant to coordinate actions around some specific goal (for example, lobbying for a reform or organising a campaign), though they did sometimes lead to initiatives of reciprocal support; rather, they were intended to help the organisations elaborate on the connections between each other's political and policy issues. Most of this work was discursive in nature, and culminated in the creation of the Causes in Common Pledge of Commitment, signed by over 70 organisations. Neither movement's perspective was left unchanged, and the reproductive justice framework transformed to include sexual justice and rights such as adult consensual sexuality and the right to gender identity and expression. In this process, the seeds of a durable collaboration were planted (Price, 2018).
Why did the solidarity between the reproductive justice and LGBTQ movements prove more encompassing and durable (transformative) than the one that took place in the context of the March for Women's Lives? Both coalitions occurred around the same period of time and in the same hostile political environment, and both the mainstream women and LGBTQ organisations were facing increased criticisms for their lack of diversity. However, the two cases differed in important aspects. Despite the inclusion of Sister Song in the march, the relation was one of suspicion, shaped by a history of marginalisation and co-optation of minority women and their issues by mainstream organisations; in the eyes of many reproductive justice activists, they were 'tainted allies' (Luna, 2010: 567; see also Price, 2010). The collaboration was also limited both in time and with regard to its scope, which was confined to the organisation of the event. Differently, the reproductive justice and LGBTQ movements shared a history of oppression and marginalisation, which fostered a greater sense of trust (Price, 2017). It was in this climate that their collaboration developed, and it benefitted from the work of entrepreneurial political actors who had shared membership in both movements (Price, 2018). Their interactions were not focused on a single issue; rather, they spanned over a number of years, with the goal of developing a common language that could make evident the connection between their issues and struggles (Price, 2017). 4 From the point of view of their transformative praxis, we also observe some differences. The four large women's organisations did not initially envision an involvement of Sister Song in the planning of the march, but conceded to its participation in decision-making and offered some resources to support their capacity as part of the negotiations that Sister Song imposed for its endorsement (Luna, 2010;Price, 2010). These efforts did not constitute a long-term commitment to the elimination of power disparities between the groups. The other coalition instead already started from a position of greater balance in terms of resources and organisation, which facilitated their coming together as equals in deliberations and development of a sense of unity and mutual respect. External donors were also fundamental in facilitating the process by providing the funding to host those events in which the intersections between the reproductive justice and the LGBTQ movements were discussed (Price, 2017).
Looking at instances of coalition work built around domestic workers' rights serves to illustrate instead the characteristics of strategies of incorporation (type 3). Given that domestic work compounds inequalities of gender, class, ethnicity/race and citizenship, it should be greatly susceptible to alliances. Nonetheless, solidarity around the rights of domestic workers is fraught with difficulties, as shown by research conducted in several countries and regions. These studies have amply documented the difficulties faced by domestic workers in achieving representation in various interest groups, such as labour unions and migrant and women's organisations (Blofield, 2012;Ferree and Roth, 1998). They also demonstrate that the isolation of these workers has often acted as an obstacle to the extension of employment and social protection to this workforce (Hellgren, 2015;van Hooren, 2017;Acciari, 2019).
Franca van Hooren's (2017) analysis offers an illuminating example of how the incorporation of domestic workers' struggles by unions in the Netherlands represented a limited form of solidarity and contributed to the persisting exclusion of domestic workers from labour and social protections. In 2006, domestic workers -largely undocumented migrants -were able to gain the support of the biggest private sector union, which was running a radical campaign on the cleaning sector at the time. The union was willing to advocate on behalf of the migrant domestic workers and lobbied other organisations (women's rights group and employers' organisations) to join in demanding the abolition of the exclusionary policy. In 2011, the campaign received new impulse by the promulgation on the International Labour Organization (ILO) Convention 189 on decent working conditions for domestic workers. However, and despite the Dutch government signing the convention, the campaign did not achieve its goal of reforming the legislation. According to van Hooren, an important reason beyond this failure relates to the frame used to campaign for domestic workers' rights.
The trade union subsumed domestic workers under the broader framework of labour rights and the inclusion of all workers in employment and social protections, thus silencing the specific needs of migrant workers. While the union's leaders justified this strategy by virtue of widespread anti-immigrant sentiments, by not naming the migrant issue, they ended up reinforcing the stereotype of domestic workers as parttime native women workers, for whom social protection was not an issue since they were already covered through the full-time employment of their partners. Thus, the use of a CD framework that incorporated the migrant domestic workers' issues under the larger rubric of workers' rights, and the prevailing conservative gender ideology in the country, contributed to maintaining the exclusionary policy. The praxis of the union was also left unchanged, and nor did it alter its structure to offer migrant domestic workers a more permanent form of representation and empowerment within the organisation.
Several studies testify that RD frames are rarely used in the struggles for domestic workers' rights (Kvist and Peterson, 2010;Spehar, 2015;Cherubini et al, 2019). Their issues are often incorporated under broader economic, human rights, workers' rights or work-life reconciliation (for the women that buy their services) frames. The hurdle of finding representation in other groups is one of the reasons why domestic workers have often resorted to building their own autonomous organisations and unions (Blofield, 2012;Hellgren, 2015;Acciari, 2019;Cherubini et al, 2019), but they still need allies to push reforms on the political agenda. Given the similarity between the issues faced by domestic workers and the devaluation of reproductive work, we would expect feminist organisations to be prone to lend their support (on this point, see also Garofalo Geymonat et al, 2021). However, several studies testify that this alliance is not so easily formed and that a number of obstacles are normally at work, including the lack of personal ties, selective issue ownership and the role that cheap domestic labour plays in allowing the emancipation of middle-and upper-class women represented in mainstream organisations (Ferree and Roth, 1998;Blofield, 2012;Spehar, 2015).
To counter their marginalisation and the limitations of strategies of incorporation, domestic workers have created their own regional platforms and transnational activist networks targeting international organisations such as the European Union (EU) and the ILO. This strategy has often led to substantial policy gains, and is also actively employed for coalition-building purposes. As stated by Acciari (2019), ILO Convention 189 is illustrative of how subaltern movements can produce a rights discourse and effectively transnationalise it.
However, transnational activism often works through pragmatic coalitions (type 4). The need to negotiate very large disparities in power, culture, strategic interests and organisational forms can induce organisations to find a solution under very broad ideas that dilute the viewpoints of less powerful partners. Schwenken's (2003) analyses of RESPECT -a European network of migrant domestic workers' and allied organisations -provide an account of this type of dynamic. The author recounts how the network actively shifted from a frame of trafficking ('modern slavery') to one accentuating women's and migrants' rights. This frame was not only more in line with the identity it wanted to construct ('the proud, empowered migrant'), but also enabled RESPECT to gain strategic access to feminist EU bureaucracies and trade unions. This strategy shows signs of a transformative praxis but is also not without its challenges. Schwenken (2003: 13) describes one such challenge faced by RESPECT in dealing with the issue of violence against women. The gender framing of the issue did not allow recognition of the differences between intimate partner violence and violence between employers and employees, and the particular nature that this phenomenon takes for undocumented migrant women working in private homes and depending on their employers for the regularisation of their immigration status. The gender framing of the issue offered limited resonance for this problem and for the need to change migration regulations. Thus, despite improvements in the representation of migrant domestic workers at the EU level and the support of feminist and class-based actors (also in gaining access to funding), the prevalent frames still needed to be transformed to make visible the specific issues faced by migrant domestic workers. The weaknesses of pragmatic forms of solidarity have been amply documented in studies of international advocacy networks and events, such as the World Social Forums (WSFs) (Vargas, 2005;Conway, 2012;Keck and Sikkink, 1998). The WSFs represented an exemplary occasion for social movements to experiment with practices to promote alliances across social justice groups. A great deal of effort was put into trying to create an open space, using an intersectional framework to increase deliberations (for example, through 'inter-movement dialogues') and the representation of groups from disadvantaged locations and groups. However, as Vargas (2005) recounts, the WSFs also remained a 'terrain of dispute, against unitary mindsets, against hegemonic imbalances and against exclusionary tendencies'.

Discussion and conclusions
Our aim in this article has been to unpack the concept of intersectional solidarity and to contribute to debates about the nature of deeply transformative forms of solidarity across social divisions. By focusing on the interplay between discursive and material dimensions of intersectional solidarity, we identified four ideal types (transformative, instrumental, pragmatic and incorporation), each with different consequences for the nature, outcomes and longevity of cooperation. This conceptualisation overcomes the opposition between 'good' and 'bad' solidarity by showing that there are several intermediate forms and that the differences between them lie in the way power disparities are recognised, represented and transformed.
Our typology also shows that strategies that unilaterally act on either the framing of issues, the sharing of resources or the forms of deliberation and representation fall short of producing an advanced form of solidarity. These aspects are clearly imbricated, but each is also irreducible to the others. The illustrations we provided clarify this point, and indeed show that using the same frame of reproductive justice was not enough to produce a durable form of solidarity when the disparities in representation and resources were not deeply transformed (Luna, 2010). Conversely, the case of migrant domestic workers' transnational activism in the EU clearly demonstrates that strategies that only focus on altering the structure of representation and funding run the risk of maintaining the issues faced by less powerful groups in obscurity (Schwenken, 2003).
Transformative forms of solidarity require trust and time; they are developed in long-term conversations in which organisations are enabled to develop a common language and make connections between each other's goals and struggles (Price, 2017). These processes are often diffused, in the sense that they are not tied to specific goals, and are favoured by the availability of resources and the presence of individuals or organisations that act as translators and bridge builders (Montoya and Galvez Seminario, 2020; see also Townsend-Bell in this issue). This work is fraught with disputes and resistances, the overcoming of which can lead to durable forms of solidarity that come close to feminist theorising about the good life (Rai, 2018).
At the core of this work is also a normative concern with regard to the extent that organisations can avoid engaging in zero-sum games and 'oppression Olympics'. Two visions of coalitions have been prominent in the literature: as sites of solidarity essential for progressive social change; and as tools of oppression in which more powerful groups get to dictate strategies and goals to the members. In practice, these are not alternative visions and often coexist in reality. Here, we have tried to speculate on the characteristics that those more advanced forms of political intersectionality -what we call, transformative solidarity -should possess. Transformative solidarity is extremely demanding -it requires the relentless work of interrogation and transformation of the material and discursive dimensions of power -and difficult to achieve in practice. As an ideal, it serves to point to the normative dimension of intersectionality, which goes beyond the temporary nature of strategic collaborations and highlights the importance of a deep and sustained engagement with the 'other' (Lyshaug, 2006).
To conclude, we would like to offer some directions to put this typology at work in future research. We believe that both a systematic meta-analysis of existing scholarship and empirical case studies would be valuable. The former would contribute to a more comprehensive understanding of both the conditions and outcomes of different forms of solidarity, while the latter would help refine the dimensions in terms of the characteristics of the frames and framing processes shaping intersectional solidarity, as well as the practices and strategies employed in interactions between actors and organisations located at different intersections of inequalities.

Notes
1 The popularisation of intersectionality raises issues about the desirability of intersectional theories and practices that do not include race centrally in their analysis (Collins, 2015;Nash, 2008), especially given the under-representation of women of colour in Western academic institutions (Mügge et al, 2018). 2 With regard to the human rights frame, Hewitt (2011: 83) states: 'the rights-based frames provide a very big tent, enabling many movement actors to join in the use of common language and still feel that their priorities are receiving attention and being validated'. 3 We thank the second reviewer for making this point. 4 Roberts and Jesudason (2013) highlight a similar dynamic in their study of the creation of a 'difficult' coalition around genetic technologies, which included disability rights advocates, women's organisations and reproductive justice activists (for a similar role played by entrepreneurial actors, see also von Wahl, this issue).

Funding
This research has received funding from the European Union's Horizon 2020 programme under Marie Skłodowska-Curie grant agreement No. 746168 -AGenDA. Any dissemination of results reflects only the authors' view and the European Commission is not responsible for any use that may be made of the information it contains.