Migrants, third union dilemma and organizational responses to migrants’ incorporation within union structures. The case study of an Italian trade union

The object of the present article is the ‘third dilemma’ of equal versus special treatment and its organizational dimension, arising in the relationship between the Union and migrant workers. In particular, the author attempts to contribute to a better understanding of how the union incorporates migrants within its own structures. Drawing on two different strands of literature on workers’ unions (one focusing on the


Introduction
The present article aims to understand how the third union dilemmaequal versus special treatment (Penninx and Roosblad, 2000) -in an organizational dimension arises within workers' unions.An important literature strand focusing on the relationship between unions and migrants explores the possible responses of the unions when they are faced with the emergence of new social subjects (Holgate, 2011;Savage, 2006;Milkman, 2000;Osterman, 2006).Within this literature strand all authors concur that the organizing strategies of migrant workers are generally taken as an opportunity for the revitalization of the unions (Milkman, 2011;Fitzgerald and Hardy, 2010).In addition to this fairly recent literature focusing on organizing strategies, it is also possible to identify another literature strand focusing on the longrunning changes of the relationship between trade unions and migrants.A pioneering work in this direction, from the early '70s, is Castles and Kosack (1973).In a very different historical context, Penninx and Roosblad (2000) identified three dilemmas unions had to face in dealing with migrant labour over a period spanning more than three decades, from 1960 to 1993.First of all, unions needed to choose whether to resist and contrast immigration or to collaborate in the immigration policy-making.Such a dilemma originally presented itself as a choice between resisting or cooperating in the policies of migrant workers recruitment enforced by governments at a time when West European countries used to have rather open immigration policies (although allowing only for a temporary residence).Yet the nature of this first union dilemma has changed over the last 50 years in parallel with the changes of the migration processes, the stabilization of migrants in the host societies and the establishment of more and more restrictive policies of admission and residence.Therefore, since the mid-1970s, the first dilemma of co-operation or resistance has always been intertwined with a second dilemma of inclusion or exclusion of migrants in the union.
The second dilemma arises when migrants have settled already in the host country and they are integrated into the labor market: that is when unions need to choose between recruiting and so including migrants, or excluding them.Also the nature of this second dilemma has undergone changes over time.In a first phase, this dilemma basically consisted in either regarding migrant workers as an integral part of the union movement ranks and thereby recruiting them as full members, or regarding them just as temporary workers and so excluding them, partially or totally, from union membership.Since the mid-1970, when the stabilization of migrants in the host country first became evident and breakthrough migrant organizations emerged to deal with the issue of the position of migrants in the labour market, the second dilemma has come to take the form of a choice between cooperating with such organizations or dismissing them as a threat to the unity of the labour movement.
Only if they include and organize migrants then, unions are faced with a third dilemma.Unions cope with the choice of whether to represent the common interests of its membership base as a whole or, rather, to represent the distinct interests and needs of migrant workers, by creating specific measures and policies and by facilitating the self-organization of migrants as special groups within the union.This dilemma has everything to do with the way in which a workers' union recognizes the identity of migrant workers.If the migratory background and geo-cultural origin of migrants are emphasized, the union may consider migrant workers as holders of special needs, interests and competencies.On the contrary, if the migrants' class interests are put in the foreground, the union may then regard the migrant worker just as any other indigenous worker.
As written above, since the 1970s, along with the socioeconomic and political changes that have affected Western European countries, there has been a deep transformation of the migration processes and migration policies.In the last 40 years, trade unions, both in old and in new immigration countries, have been progressively showing more openness towards migration and the recruitment of migrants.This tendency has led several scholars to consider at least the first two of the three dilemmas here identified as currently uncritical: migrants have actually reached and gained access to trade unions.This does not mean that the unionization of migrants has no problematic dimensions, but it is evident that in the recent past the relationship between unions and migrants has been structured.

Italian unions, migrants and the third union dilemma
The relationship between trade unions and migrants in Italy could be viewed as emblematic of what explained above.Some scholars speak of an extended advocacy coalition (Ambrosini, 2005) when referring to the associations and organizations that play a pro-migrant role in Italian political and social arena, and they consider Italian unions as an important part of this advocacy coalition (Ambrosini, 2013).During the last thirty years the CGIL, the CISL and the UIL, the three main Italian workers' union confederations, have often led campaigns to extend the rights of the migrants.One could argue about the actual perseverance and effectiveness of the strategies adopted by the Italian unions, but there is wide agreement that Italy has never seen a strong conflict between migrants and unions.Italian unions have in fact maintained a critical stance against restrictive and discriminatory immigration policies and they have always championed the rights of migrants.
At the same time, Italian unions have unquestionably made a substantial contribution to the integration of migrants into the Italian society, providing over the years an unsought yet invaluable and all the more necessary compensation for the shortcomings of the State (Mottura and Pinto, 1996).On the local level, unions are important (often essential) agencies for migrants' integration.For instance, all local unions are equipped with special bodies dedicated to assist and help migrants find their way through the complicated and uncertain Italian legislation and they also have activated collaborative relationships with public authorities in the field of migration management (Ambrosini, De Luca and Pozzi, 2014;Mottura, Cozzi and Rinaldini, 2010).Furthermore, since the very beginning of the migration process in Italy, the number of migrants affiliated with a workers' union has been relevant.The growth of the unionization of migrants has gone hand in hand with the growth of the presence of migrants in Italy.In 2012, about 50% of migrants in Italy with a regular employment were members of one of the three main Italian union confederations (data processing from Labour force survey by Istat and Centro Studi e Ricerche IDOS).In the same year, migrants amounted to 8% of the total union membership (Ibidem).Undocumented migrants, who have no regular employment (which is not possible to have without a valid permit to stay), cannot be recruited as union members, but they usually turn to Italian unions especially on the occasion of a mass regularization (undocumented migrants can be exceptionally enrolled as unemployed, but currently this sort of expedient is not widespread and, above all, it is a discretional practice).
Despite the three union dilemmas cannot be solved once and for all, there is wide consensus that the first two dilemmas can be considered as currently uncritical.At the same time, several scholars point out that the third dilemma -equal versus special treatment -is a current challenge for the union (Connolly, Marino and Martinez Lucio, 2014;Marino 2012).In particular, migrants' inclusion and their integration within union structures is a critical issue.It is no coincidence that Penninx and Roosblad underline just this issue by describing the third dilemma: «special policies may also pertain to the integration of immigrant (…) into the organizational structure of the union itself (…) At one extreme there is the minimal variant that creates facilities for immigrant workers to organize themselves as special groups within the union, but outside the existing core organization, for example as subsidiary, often advisory bodies.At the other end of the scale are unions that have become conscious of the culturally and ethnically bound nature of their organization and the barriers that immigrants have to surmount to participate fully and equally.They may implement more drastic specific measures to adapt that organization to the new composition of its membership and try to guarantee the influence of the new membership in decision making.Such unions may launch active anti-discrimination policies within their own ranks (…) implement positive action in training and employing immigrants (…) within their own ranks, and in getting there more strongly represented in the higher decision-making ranks of the unions.It may be assumed that such policies will meet with much more resistance» (Penninx and Roosblad, 2000: 11-12).
Once again, the Italian case can be considered emblematic.In the process of collective bargaining negotiations, unions show a tendency to waver constantly between the recognition of the geo-cultural specificities of the migrants and equal treatment.In 2003 about 10% of the national industry collective agreements contained explicit references to migrant workers, but from several researches carried out on the subject what emerges is the scarcity of their effective implementation on the workplace level and the lack of a coherent union strategy (Mannoia and Salieri, 2007).It is no coincidence that bargaining agreements at company level usually contain very few explicit references to migrant workers, inasmuch as the formal process of collective bargaining on the local level tends to be balanced by informal negotiations.In general, a mixed approach to migrants issues is prevalent within single company agreements, that is a sort of combination of both universalistic and particularistic approaches.(De Sario, 2014).
As regards the organizational level, it is widely known that the main Italian union confederations are equipped with special bodies dedicated to migrant workers: since the 1980s, special bodies called Centri Lavoratori Stranieri (CLS) have been created within each Camera del Lavoro (the local structure of the CGIL); at the same time, the CISL has supported the establishment of the Associazione Nazionale Oltre le Frontiere (ANOLF).A CLS and the ANOLF differ in several aspects.The main difference is that the former is a special body internal to the union organization, the latter is an external association promoted by the union confederation and linked to it by statute.This difference implies other important peculiarities concerning the ways in which migrants join the workers' union and the very nature of their membership.(Mottura and Pinto, 1996).Nevertheless, the creation, development and consolidation of both ANOLF and the CLSs has always been characterized by an unsolved tension (emerging unmistakably from the discussions that animate the periodic congresses of the two main union confederations) between universalistic pressures to move beyond the notion of a special body (asking thus for its dissolution and integration within the union organization) and particularistic approaches that consider the specific needs of migrant workers and support their permanence.In addition to these special bodies, each union confederation has created special committees and teams with the aim of designing union strategies on migration issues.With respect to migrants' inclusion and integration within the union structures, there is no doubt that the number of migrant union representatives in the workplace and the number of migrant union officers has increased over the years.However, the process of inclusion of migrants does not appear to be a smooth one nor are migrants' career paths within the union unimpeded, as an existing problem of organizational segregation may seem to suggest.
To sum up, it is legitimate to argue that Italian unions have been successfully engaging in two of the three dilemmas of the relationship union/migrants identified by Penninx and Roosblad (2000).However, the third union dilemma seems more complex and it opens up various issues, including migrants' inclusion and integration within union structures.This is precisely the object of this paper.

The third dilemma in an organizational dimension: an analytical framework
An analysis of the inclusion and integration of migrants within the union as an organization requires us to focus on the logic of organizational action (Zan, 1988) governing migrants' inclusion and integration in the organizational structures, whereby the logic of organizational action is defined as the intentional and bounded rationality (Simon, 1958) that influences the behavior and performance of organizations (Zan, 1988).At the same time, such an analysis demands that we recognize that the union is a peculiar organization, a sort of 'combination of movement and organization' (Grandori, 1978).This is also the focus of several classic organizational studies on workers' unions published in the second half of the twentieth century (Lipset, Trow and Coleman, 1956;Child, Loveridge and Warner, 1973).These studies coincide in recognizing the presence, within the union, of both bureaucratic rationality (routinization of operations, specialization of functions, uni-directional communication, fast decision-making, concentrated coordination and control system, in which the primary source of authority stands at the top of the hierarchy) and representative rationality (flexibility of operations, redundancy of functions, multiplicity and multi-directional communication, slow decisionmaking, distributed coordination and control system, in which the primary source of authority lies at the base of the hierarchy).In more recent years, a vision of the union as a system has been established, in which complexity derives mainly from the "weak character" of the relations among the various organizational units (but also among activities, roles and functions) (Zan, 1992;Zan, 2011).Starting from the concept of "loosely coupled system" (Weick, 1976), the union is seen as a network of organizational ties that tend to be loose and where different logics of organizational action occur, including the administrative and the representative ones (Zan, 1992).All this allows us to consider the union as an organization formed by both bureaucratic and adhocratic structures.In brief, administrative and representative rationalities represent a transversal conceptual heritage which crosses the different theoretical perspectives through which the union as an organization has been studied.The logics of organizational action governing migrants' inclusion and integration within the union as well are necessarily innervated by these rationalities.Hence the following first argumentation is proposed.
Argument 1: a coexistence of two logics of organizational action governing migrants' inclusion and integration characterizes the union.The first logic of organizational action is oriented towards the organization (routinization of operations, specialization of functions, uni-directional communication, fast decision-making, concentrated coordination and control system, in which the primary source of authority stands at the top of the hierarchy) and the second one is oriented towards the movement (flexibility of operations, redundancy of functions, multiplicity and multi-directional communication, slow decision-making, distributed coordination and control system, in which the primary source of authority lies at the base of the hierarchy).
As stated above, the third union dilemma -equal versus special treatment -is extensively linked to the notion of migrants' identity held by the union.Connolly, Marino and Martinez Lucio (2014) identified three main logics of union action which shape union discourse and policies on migrant-related issues.The first one is the logic of class, which leads unions to consider migrants as part of the wider working class, thus sharing the same interests with the indigenous workers.The second one is the logic of ethnicity, according to which migrant workers have distinctive interests and needs that require the promotion of specific policies.The third one is the logic of social rights, which considers migrants as potential citizens and so leads the union to engage with issues that are not directly workplace-related.These logics are in continuous tension and the union implicitly or explicitly resort to all of them in the process of building representative action.The third dilemma in an organizational dimension shares the same characteristics.The logics of organizational action governing migrants' inclusion and integration within the union are inevitably linked to how the identity of migrants is considered.In particular, these logics tend to be linked to the first two notions of migrant identity mentioned above: migrants workers can be viewed just as all the other workers and when they are included within the union structures they are considered just like any other union organizer; alternatively, their migration background and geo-cultural origin can be emphasized and so their specific interests and competencies can be taken in consideration when they become union organizers.In the first case, the logic of organizational action governing migrants' inclusion and integration within the union implies the minimal organizational variants.In the second case, the logic of organizational action governing migrants' inclusion and integration lead the union to implement specific measures to adapt the organization to the new composition of its membership.Therefore the theoretical framework can be complemented in the following way: Argument 2: the logics of organizational action governing migrants' inclusion and integration within the union depend on how migrants' identity is considered.If a notion of class identity prevails, the logic of organizational action implies only minimal organizational variants.If a consideration of migration background and geo-cultural origin prevails then the logic of organizational action involves the establishment of specific organizational measures and policies.
To sum up, the logics of organizational action governing migrants' inclusion and integration are simultaneously characterized by a tension between class identity and the emphasis on migration background and geocultural origin.Although the inclusion and integration of migrants are not called into question, by intertwining this tension with the orientation of the logic of organizational action, it is possible to identify three ideal-types of organizational responses to the issue of the incorporation of migrants within the union structures.
When the logic oriented towards the organization is combined with the recognition of the migration background and the geo-cultural origin, the organizational response can be defined as a 'bureaucratic adaptation'.Drastic organizational changes are made, policies and measures for migrants are activated and special bodies and structures are established within the union.The rationality of these changes is characterized by a formalization and specialization of the functions, concentrated coordination and control, and top-down decision making.
When the logic oriented towards the movement is combined with the recognition of the migration background and the geo-cultural origin, the organizational response can be defined as an 'adhocratic adaptation'.In this case, organizational changes are made within the union, but the rationality of these changes is characterized by a flexibility of functions, informal procedures, distributed coordination and control and bottom-up decision making.
Independently from their orientation, when the logics of organizational action are combined with the consideration of class identity, the organizational response can be defined as 'unresponsive reception'.Minimal organizational variants are established.
It is important to underline that all these response are always present in the union and that under any circumstances the migrants' inclusion and integration is eventually reached.

Aim and methodology of the research
The research exposed in the following pages was carried out using a qualitative methodology.Its specific aim is to understand the characteristics of the organizational responses which govern migrants' inclusion and integration within the CGIL.
Consistently with the adopted theoretical framework, the characteristics of the organizational responses are to be found in the discursive practices of the actors that live within the organization and contribute to structure it, as well as in the meanings that the actors attach to these practices.
In the following pages, the results of two connected research experiences are presented.Semi-structured face-to-face interviews were conducted with union organizers of the CGIL of Emilia Romagna (an area in northern Italy with a high presence of migrants) at different hierarchical levels.The first research (2008 -2010) involved 40 Italian union organizers of four CGIL federations (FILCAMS-CGIL, service industry; FLAI-CGIL, agri-food industry; FIOM-CGIL, engineering and metalworkers; FILLEA-CGIL, construction industry); the second research (2010 -2012) involved 69 migrant union organizers of CGIL (mainly union representatives at workplace level) of the same four CGIL federations.These interviews were added to the interviews with the leaders of the main Camere del Lavoro of Emilia Romagna and the CLS organizers.
The interviews addressed several aspects of the relationship between the workers' union and the migrants, but they focused mainly on a specific core issue: the ways in which the trade union recruits, includes and integrates migrants within its own structures.The interviews also provided an opportunity to collect a considerable amount of instructive material about migration-related issues (publications, union internal documents, internal written communications, internal directives, etc …) and the time dedicated to the research presented several opportunities to attend at internal meetings, conferences, congresses, committees and decision-making bodies of CGIL, whenever migration-related issues were debated.
The research design has facilitated the construction of a heterogeneous sample and an articulated and complex textual corpus.This textual corpus has been analyzed according to a soft version of Grounded Theory (Glaser and Strauss, 1967).

Logics of organizational action governing migrants' incorporation within the CGIL structures: The shift from an "unresponsive reception" to an organizational adaptation
The relationship between migrants and the CGIL is not recent.The first migrant members in the CGIL date back to the mid-1980s.Several studies on the relationship between migrants and the CGIL identify a clear break point in the early 1990s (Basso, 2004;Mottura and Pinto, 1996;Tassinari, 2013).Up to that period, all migrant-oriented actions on part of the CGIL were aimed at providing general assistance services through the organizational structures called Comitati di Informazione per Disoccupati (CID, Committees for Information to the Unemployed) which, by the way, were not established with the idea of meeting the migrants' specific needs.Migrants in fact were then regarded simply as the weakest subjects in the labour market and the CGIL did not recognize them as holders of specific interest or needs.The CGIL would provide assistance, support and information activities to the migrants, but no specific organizational structures had been established up to then.That was also when the incorporation of migrants as union organizers started, but on the basis of the theoretical framework described above, in this first period it is possible to recognize the prevalence of a logic of organizational action governing migrants' inclusion and integration combined with the consideration of class identity.The organizational response tended to be that of an "unresponsive reception".
In the early '90s, the CGIL constituted special bodies called Centri Lavoratori Stranieri (CLS).The CGIL gradually became conscious of the specific needs and interests of migrant workers and it equipped itself to deal with these specificities.This organizational change formed part of a more general process of reconsideration of union action on migration-related issues.This period of self-reflection ended in 1996, with the XIII Congress, when the CGIL understood the need to move beyond its role of compensator for the shortcomings of the central and local public authorities on matters of migration.This marked the start of a new phase for the CGIL, one in which its relationship with migrants became increasingly far more oriented towards the dimension of collective bargaining and union protection (while still keeping the union's role of support and information provider).
In order to provide services and support to migrants, in a matter of a few years each Camera del Lavoro established its own CLS.Over time, the CLSs became important special bodies providing a number of various services (assistance, support and information about different issues, including the migration law) specifically targeted to migrants.Simultaneously, other union structures (including the union federations) should have taken the responsibility to represent the migrant workers and to deal with migration-related issues.The CLSs should have taken in charge the task of coordinating the actions of the different union structures on migration-related issues, instead of taking on an exclusive representative role for migrant workers.All this was supposed to generate both the process of diffusion of responsibilities on migration issues within the union structures and the integration process of competencies and skills.Actually, over 20 years the CLSs have played an important role in building strong relationships between the CGIL and migrants.Migrant members in CGIL federations has kept on growing and nowadays migrants are an important component of the total union membership.However, the transition described above brought to light several ambiguities.The CLSs do not have a formal representative role, rather, they tend to be considered as the specific union structures in charge of dealing with migration issues; in this perspective the task of reaching out to migrants rests with the CLSs only, which as a result relieve all the other union structures of this responsibility.A reorganization of the CLSs aimed at reinforcing their integration with the other union structures was recently started, but the results are still uncertain (Danesh, 2014).Moreover, although the process of migrants' inclusion and integration within the union structures has strengthened, the modalities of inclusion and integration have often been characterized by opposite pressures: on the one hand, migrants have been mainly expected to carry out activities targeted to migrant workers; on the other, migrants' commitment within the union has always been expected to be no different from that of any other union representative (Mottura, Cozzi and Rinaldini, 2010).In other terms, over time, in parallel with the tension to incorporate migrants just as any other workers, pressures to recognize the specific needs and interests of the migrants emerged within the CGIL and several organizational changes followed.The logics of organizational action governing migrants' inclusion and integration have gradually become ever more intertwined with a consideration of the migration background and the geo-cultural origin, but all this came to structure a complex organizational context.Indeed, the analysis of the textual corpus produced by the research brings distinctly to light the current coexistence of two organizational responses to migrants' inclusion and integration: the "bureaucratic adaptation" and the "adhocratic adaptation".

The "bureaucratic adaptation" as an organizational response to migrants' inclusion and integration within the CGIL
The considerations arising from the views on migrant union organizers' identity expressed by some interviewees seem point to two specific dimensions of identity: the geo-cultural origin and the migration experience.Geo-cultural aspects appear to differ substantially, depending on the interviewee, but in general, all arguments tend to reflect a pars pro toto view.The migration experience also tends to be seen as a sort of undifferentiated experience.Given such meanings of identity as it is attached to migrant union organizers, it is understandable that specific resources are identified in relation to the geo-cultural origin and the migration background: cultural mediation skills (knowledge of foreign languages, knowledge of the origin culture, etc ...) and the ability to deal with migration issues.
These interviewees argue that the CGIL should include migrants through "dynamics of co-optation", that is a direct selection based on the recognition of the migrants' specific attitudes, knowledge and skills.In the same perspective, the interviewees are in favor of the establishment of quotas dedicated to migrants within the union structures.In this group of interviewees there is a widespread belief that quotas are a necessary evil to implement migrants' inclusion.
«When the union co-opts the worker to put him inside the organization, this can be a problem.The best way is legitimation from below (...) On the other hand it is unlikely that this will happen now (...) If we wait for these migrants to be elected as union representatives, it will never happen.So we need to promote migrants» (union organizer, man, Italian, 29).
All interviewees recognize that there are some critical aspects on the organizational level of the union, like the separate career paths and the organizational segregation of migrants working in the union.At the organizational meso-level, the CLSs are the union structures where these criticalities come to light.However, according to these interviewees, keeping the CLSs functioning is a priority because their specific function always has ensured and keeps ensuring the accumulation of an expertise and of distinctive competencies which should be safeguarded.At the same time, the specialization within each union structure of the union activities, as these are specifically targeted to migrants, is considered a necessary organizational response to the specific working lives of the migrants.So the problem of organizational integration is not approached in the terms of a move beyond the existing special bodies within the union and the related activities specifically targeted to migrants; rather, it is addressed in the terms of a reallocation of the union personnel that is employed in the same structures and activities.This involves the introduction of binding rules imposing limits to the number of migrant union officers employed within the special bodies and in the activities that are specifically targeted to migrants.
«In union federations (...) if you let it be spontaneously, what happens is that migrant union officers (…) turn up dealing with migrant workers (...) you have to make sure that all workers turn to them (…) and that Italian union officers take care of migrants» (union organizer, man, Italian, 32).
To sum up, what emerges is that all interviewees in this group argue that migrants' incorporation within the union structures should be based on the identification of the specific migrants' resources and should be supported by innovative organizational measures, such as the establishment of quotas and dynamics of co-optation and reallocation of union personnel.All these measures can be ascribed to a "bureaucratic response" to the issue of migrants' inclusion and integration.Indeed, according to the interviewees, these same innovative organizational measures can be implemented through the tendency to reinforce a typical bureaucratic rationality that is characterized by a formalization and a specialization of the functions, fast decision-making, coordination and concentrated control system, with the main source of authority at the top of the hierarchy.

The "adhocratic adaptation" as an organizational response to migrants' inclusion and integration
Even though all the interviewees acknowledge the importance of the geo-cultural origin and the experience of migration, some interviewees tend to dismiss the idea of limiting the identity of migrant union organizers to these two dimensions.In any case, according to these interviewees, the geo-cultural origin appears to be much more multi-faceted and the migration background is taken as a dynamical and differentiated process.
«I don't understand (ironically) who this migrant with a capital M is (…) Migrants are women men, young and elder, they are from Bosnia, Senegal, China … they are married and single (…) They came here for different reasons and they have different projects (…) They are good people or dickheads (…) I don't mean to deny that there are diversities and if we [the CGIL] are not able to integrate these diversities, this means we have a big problem, but it is all too easy to create the two categories "migrants versus not migrants"» (union organizer, man, Italian, 6).This does not amount to downplaying the differences (in a way, differences appear more emphasized and stressed in this perspective), neither to a contraposition between class identity and a recognition of the migration background and geo-cultural origin; yet the interviewees are inclined to argue that the identity cannot be ascribed by the organization on the basis of preconceptions.
All this, though, seems to limit the possibility for the union to accurately identify the diversity of migrant union organizers and their specific resources and competencies.It is not a coincidence that these interviewees argue that the CGIL should aim to include migrants within the union structures through a "dynamic of democratic validation" in the workplace.
The election at workplace level should constitute a test of legitimacy that is essential for migrants and Italians alike and just this should allow for both the inclusion and the integration of migrants within the union structures.To be elected as a union representative means to represent all workers and thus the "dynamic of democratic validation" safeguards against the separation of the career paths and the organizational segregation of migrants working in the union.
«If a migrant worker is elected (…) he is a union organizer (...) nobody can tell him to work for migrants only, to go to the CLSs, nobody can force him» (union organizer, man, migrant, 16).
In this perspective, the interaction with colleagues at workplace level with a view to meeting their approval represents the main training opportunity both for migrant and Italian future union organizers, who need to interact and exchange views at workplace level, if they wish to integrate differences.Therefore both at meso-and micro-organizational level the integration should be a relapse of the "dynamic of democratic validation".«The migrants' inclusion happens with superficiality (…) Often we think that it is enough to appoint some migrant union officers to work in the CLS or within a union managing board of the Camera del Lavoro or within some union federation managing board … we think that this is enough to become conscious of migration issues (…) but this doesn't work, it is not working (…) if we really want to build a new generation of migrant and Italian union officers able to represent all the workers, we have to establish paths by putting them on probation through discussions, learning and agreements and such path is to be found at workplace level» (union organizer, man, Italian, 31).
Obviously the "dynamics of democratic validation" is not perceived as uncritical.In particular, these interviewees acknowledge that the risk that the vote can be strongly influenced by the geo-cultural origin of the candidates rather than by their union competences is out there, but they also argue that the legitimacy conferred by the vote cannot be replaced by co-optation or quotas.The quotas within the union structures at different levels, in this sense, are not meant as a tool, but rather as a goal.
«Exceptional cases?Why not?There is no rules (…) A migrant can go straight into the union without … being elected as representative in the work place, but he takes a risk, because he will be always perceived as "migrant" (…) The election is important because it proves that you can represent everyone» (union organizer, man, migrant, 19).
According to the interviewees, the innovative measures have to be aimed at facilitating the interaction and communication on the workplace level, in order to diffuse and share competencies and knowledge among union organizers, union representatives and workers.In this sense, the most incisive measure is the design and implementation of learning paths aimed both at Italians and migrants, which can involve both union organizers and workers.
«We have already done a lot of things for migrant workers (…) but I think that it is important to focus on union training for them [migrant workers] (…) of course we have also to create informed Italian workers.We have to find the way to learn from each other» (union organizer, man, Italian, 35) «Migrant union organizers are holders of new skills (…) but we have to build learning opportunities to transmit and share these competencies and we are working in this direction» (union organizer,man,Italian,22) What characterizes this perspective is that, in order to ensure migrants' integration and to avoid their organizational segregation, the organizational changes considered as the most incisive are located at the bottom of the union, near the borders of the organizational field, where it is most evident the exchange between the union and its environment.In other words, action during the phase that precedes inclusion appears to be crucial for migrants' incorporation within the union structures.Organizational tools like quotas, dynamics of co-optation and reallocation of union personnel are not viewed as an innovative organizational measure and, above all, they are not considered an effective one.Facilitating the dynamics of democratic validation, supporting workers' interactions on workplace level and designing learning paths in order to share competencies and knowledge are the crucial issues for migrants' inclusion and integration within the union structures.These measures seem to be ascribable to an "adhocratic response" to the issue of inclusion and integration of the migrants.Indeed these organizational measures can be implemented through the tendency to reinforce a typical representative rationality that is characterized by a flexibility of functions and procedures, a distributed coordination and control, with the main source of authority lying at the base of the hierarchy.

Discussion
The general aim of this article consists in an attempt to contribute to a better understanding of how the third union dilemma in an organizational dimension arises within workers' unions.Drawing on two different strands of literature on workers' unions (the first on the union as an organization; the second on the relationship between migrants and unions) a theoretical framework was generated and three ideal-types of organizational responses were identified: the "bureaucratic response", which weaves together the tension to consider the workers' geo-cultural origin with the logic of organizational action oriented towards organization; the "adhocratic response", which weaves together the tension to consider the workers' geo-cultural origin and migration background with the logic of organizational action oriented towards movement; and the "unresponsive reception", which weaves together the logics of organizational actions, independently from their orientation, with the tension to consider the class identity.At a later stage, by using this theoretical framework, the migrants' inclusion and integration within an important Italian trade union, the CGIL, was explored.
The qualitative research here exposed brings to light the characteristics of the organizational responses which govern migrants' inclusion and integration within the CGIL structures.What emerges from the analysis is that the CGIL gradually shifted from an "unresponsive reception" to an organizational adaptation.This obviously does not mean that the CGIL rejected the class identity and the respective logics of organizational action, yet there has been a growing pressure to recognize the specific needs and interests of the migrants and several organizational changes followed within the CGIL.
All this structured a complex organizational context where two organizational responses to migrants' incorporation coexist: the "bureaucratic adaptation" and the "adhocratic adaptation".
Despite both responses aim to incorporate migrants within the union structures and both also aim to deal with the organizational problems arising within the union (for instance, the organizational segregation of migrants working in the union), these two responses involve different organizational measures and policies and they can also come into conflict.Indeed, on the one hand, the 'bureaucratic response' involves the establishment of quotas, the co-optation of migrants and the reallocation of union personnel, while on the other hand, the 'adhocratic response' involves the reinforcement of the dynamics of democratic validation, the support of workers' interactions on workplace level and the design of learning paths aimed at sharing competencies and knowledge.The modalities through which migrant workers are included and integrated within the CGIL structures appear to embedded within these tensions.
The aim of the research was not to understand which organizational response is prevalent within the CGIL, neither which response effectively proves to facilitate and is more efficacious for the sake of migrants' incorporation.These questions can represent future research trajectories and it would be interesting to implement such trajectories in a comparative dimension.Nevertheless, maybe the more interesting issue is not the predominance, but rather the balance between the organizational responses.Maybe the strong predominance of only one of the organizational responses may limit the opportunities for migrants' inclusion and integration.Up to the present, the coexistence of these different organizational responses and the way in which they qualify the third union dilemma surely has not precluded migrants' incorporation within the CGIL.A case can be made that the union should not try to solve the tension generated by the third dilemma in an organizational dimension; it should rather preserve and treasure it.