Institutionalizing Precariousness? The Politics of Boundary Work in Legalizing Agency Work in Germany,1949–2004

Strategic action fields

For deriving a heuristic analysis of the politics of boundary work in such a setting, here, the notion of boundary work as deployed in Zietsma and Lawrence (2010) is cross-fertilized with the idea of inter-field relations from the theory of strategic action fields (Fligstein & McAdam, 2011, 2012). Examining organizational fields through the analytic lens of strategic action fields allows making long-term processes of political contestation visible. Fligstein and McAdam (2012, p. 9) define a “strategic action field” as a “constructed meso-level social order” that can form around policy issues and organizational practices. These orders typically consist of three elementary actor groups with varying resource endowments: incumbents dominating the field, challengers striving to change a field’s order, and governance units that regulate the field. Incumbents and challengers diverge in orientation and goals due to competing frames of reference, and both groups try to achieve their purposes through collective action. With collective action, Fligstein and McAdam explain what makes conflict endure, even become institutionalized as Dahrendorf (1959) argued, as challenging groups appropriate or create in collective action their organizational vehicles for sustained contestation (Fligstein & McAdam, 2012, p. 21).

Clearly, not all organizational fields are best described as being a strategic action field as there might be fields with either more unified or more opaque governance structures such as those forming around uncontroversial organizational practices or within communities respectively (see Thornton, Ocasio, & Lounsbury, 2012, p. 164). Nonetheless, it is justified to apply the Fligstein and McAdam (2012) lens to those fields forming around “bread and butter” issues that involve the state and the law and are the scene of struggles between deeply rooted but conflictive social expectations on organizations (or political division, for that matter, e.g. Kriesi, 2010). Institutional dynamics in these fields gain a controversial edge when bringing about (new) formal lawmaking as authoritative structuration (Giddens, 1984), which imposes social categories as boundaries between social groups, with profound effects for individuals’ life chances (Lamont & Molnár, 2002; Tilly, 1999, 2004). This is regularly the case in contests about labor law (Deakin, 2006) which was – among others – noted by the late Max Weber:

The increasing significance of freedom of contract … implies a relative reduction of that kind of coercion which results from the threat of mandatory and prohibitory norms. … But it is also obvious how advantageous this state of affairs is to those who are economically in the position to make use of the empowerments. (Weber, 1968, p. 730)

The controversial edge is amplified when the regulatory contest is connected to social movements concerned with organizational practices (Schneiberg & Lounsbury, 2008) inasmuch as practices embody a field’s regulatory settlement, i.e. the formal rules applicable beyond a single organization (Vaara & Whittington, 2012, p. 312). Therefore, organizations become arenas of law enforcement, making organizations inclined to politically influence lawmaking outcomes (Edelman & Suchman, 1997; Hargrave & Van de Ven, 2006; King & Pearce, 2010); even up to the point to strive for hegemony (Levy & Scully, 2007) in defining a field’s settlement.

Boundary work

Fligstein and McAdam (2011, 2012) insist that strategic action fields are structurally embedded in other fields, i.e. have ties to (a set of) neighboring fields, and the history of a focal field is influenced by developments in these fields, sometimes even dominated by them. Similar to approaches theorizing the relative permeability of fields to outside influences as conducive for radical change (Greenwood & Hinings, 1996), Fligstein and McAdam give priority to the structural properties of inter-field relationships to explain how inter-field relations contribute to changes within a given field as these convey and mediate exogenous shocks (Fligstein & McAdam, 2011, p. 8). For example, the authors see a high number of ties or a high interaction frequency between fields as conducive to the spread of exogenous shocks shaking a focal field’s order.

As a complement to that the theory of institutional work provides a more detailed view on the agentic part of “field structuration” (DiMaggio & Powell, 1983, p. 148), i.e. how a field’s membership, governance, purposes and cognitive understandings are defined (Suddaby et al., 2007; Zietsma & Lawrence, 2010). Opening up to a practice-based perspective (Barley & Tolbert, 1997; Nicolini, 2012), the politics of boundary work can be conceived as concerned with symbolically constructing and practically enacting boundaries (e.g. Abbott, 1995; Lamont & Molnár, 2002). Boundary work can mean a myriad of different practices such as creating, expanding, reinforcing, crossing, blurring, opening, permeating, undermining or disrupting social boundaries (e.g. Gieryn, 1983; Llewellyn, 1998) between groups, organizations and fields across time periods, spatial orderings or institutional levels. For example, “boundary-work reproduces the organization as an ordered entity” (Llewellyn, 1998, p. 23), but it is also the vehicle for organizational change as organizational boundaries are redrawn, crossed and permeated in attempts to manage the tensions at organizations’ internal-external interfaces (Gawer & Phillips, 2013).

Here, I concentrate on boundary work in and across organizational fields as redrawing the brackets around organizational practices by structuring a field from within, as well as using relations to neighboring fields to create (new) insides and outsides of the focal field, a process that Abbott (1995, p. 872) calls “linking up” or “yoking”. Hence, I understand the politics of boundary work as dealing with how a field’s boundaries are manipulated to attack or defend a field’s settlement for shaping organizational practice. Boundary work within fields, then, is concerned with maintaining or changing a field by manipulating its boundaries from the inside. Boundary work across fields aims to shield or lever a given field settlement by relating the focal field to neighboring or higher-order fields.

Boundary work within strategic action fields

Boundary work across strategic action fields

Institutional dynamics

For institutional dynamics, the politics of boundary work make a field’s settlement contestable, even if it is firmly protected by incumbent coalitions and enshrined within legal rule-setting. This is because contestation and haggling between incumbents and challengers also go beyond a field’s boundaries, making a field’s regulatory status quo into a precariously reproduced settlement in itself (Hargrave & Van de Ven, 2006; Seo & Creed, 2002; Thelen, 2009). In the long term, there is even the possibility of what Fligstein and McAdam (2011, p. 17) call “genuine transformation”, i.e. the overturning of a field’s order by the elevation of the challengers to a position of dominance within the field, making (former) challengers the “new kings”. Over time, then, fields might move back and forth between divergent settlements on regulating organizational practice, depending on the (re)formation of a field’s external boundaries and on the alternation of power positions brought about by field participants’ “jockeying for position” (Fligstein & McAdam, 2011, p. 6). However, how institutional dynamics are modulated into (recurrent) contests about organizational practice by the politics of boundary work is primarily an empirical question about (1) what practices of boundary work within and across fields are actually used by incumbents and challengers respectively, and (2) how these practices influence a focal field’s settlement.

Data and methods

To investigate how this regulatory change has been affected by the politics of boundary work across and within fields over time, process-oriented data collection and data analysis are required (Langley, 1999; Lawrence et al., 2001) to capture multiple perspectives on that process (McAdam, Tarrow, & Tilly, 2008). Therefore, a variety of data sources is used (see Table 1). For each field, documents have been collected for analyzing regulatory outcomes (e.g. ILO conventions, German Temporary Agency Work Act, collective agreements) and processes (e.g. press releases, campaign materials, industry reports, companies’ annual reports, and publicly available minutes) over time. Conducting a historical case study, these documents are the “primary” source, as only these documents reveal how boundary work across fields has contributed to the long-term changes in the domestic field (see Appendix A for a detailed listing). The documents not only check for informants’ retrospective or interest biases, but are themselves sources regarding the field’s structure, earlier boundary work, and the role played by the field’s governance units.

OPEN IN VIEWER

Table 1.

Data sources.

Source type		Domestic field	Transnational field	Total
Document analysis	Formal documents on regulatory outcomes	e.g. German Law on Temporary Agency Work (AÜG), Collective agreements, Federal Court decisions	e.g. ILO conventions & recommendations, European Union directive, Joint declarations, European Court of Justice decisions	21
	Actors’ documents	e.g. Federal government reports, employer associations’ and trade unions’ documents, archive material, historical documents	e.g. ILO reports, Company documents, Employer associations’ and trade unions’ documents, official conference protocols	32
Semi-structured interviews with	Actors directly involved in negotiations on agency work (unions/employers)	12/6	5/3	26
	Actors involved in negotiations on agency work from a client industry perspective (unions/employers)	6/4	5/–	15
	Agency industry (labor representatives/management)	3/7	1/1	12
	Client companies (labor representatives/management)	7/7	−	14
	Sum	28/24	11/4	67
Additional sources	Other documents	secondary quantitative data sources, press articles		7
	Public media interviews	4	−	4
	Experts in international labor law	−	2	2

In order to obtain a more complete picture of the nature of boundary work, 67 self-conducted interviews were used with labor (39) and business representatives (28) from the domestic and the transnational field, stemming from two related research projects between 2009 and 2013. All respondents have privileged insight into the field of agency work from various points of view (client firms, agencies, employer associations, worker representatives). Due to their executive management functions and responsibilities as CEOs, employer associations’ and union officials, HR managers or works council heads, many act as boundary spanners, negotiating and coordinating between labor, business and state agencies (Perry & Angle, 1979); and a dozen had been personally involved in key events between 1997 and 2004. Although all these interviews contain retrospective reflections on the field’s historical development to varying degrees, they give a more detailed picture of the boundary work at the time they were conducted. The interviews were on average 40 to 50 minutes long, the shortest under 30, the longest about 80 minutes. The interviews were conducted by one or two researchers, 35 being face-to-face and the rest by telephone. With three exceptions, the interviews could be audio-taped and were transcribed.

Analytic process

For data analysis, thematic coding (Flick, 2006) was applied and carried out in the following steps:

OPEN IN VIEWER

Figure 3.

Data structure.

Incumbents activating, upholding and reinforcing boundaries

Incumbents in the agency work field engaged in three types of boundary work in order to maintain the field’s settlement: activating boundaries, upholding boundaries and reinforcing boundaries ideologically.

Incumbents protecting external boundaries

As far as incumbents engaged in protecting the field’s external boundaries, in the field of agency work, incumbents made use of two practices: reaching out to neighboring fields and reaching out to higher-order fields. A good example of reaching out to neighboring fields is given by the post-war field incumbents’ installing of a tripartite structure in almost all related fields such as labor law, social security, vocational training and employment policy (Paster, 2012). By reaching out to higher-order fields incumbents attempted to ensure that their policy is in line with transnational trends. For example, during negotiations about the EU directive on agency work in 2002 – finally adopted in 2008 (Countouris & Horton, 2009; EU, 2008) – the German government made suggestions in line with the new incumbent coalition’s model of regulation (Ahlberg et al., 2008; Jones, 2002).

Challengers redrawing and blurring boundaries

In the field of agency work challengers engaged in two kinds of boundary work in order to change the field’s settlement from within: challengers attempted to redraw boundaries and to weaken or to blur boundaries symbolically.

Challengers crossing boundaries

For challengers in the field of agency work, the crossing of boundaries in order to change the field’s settlement was decisively important; their repertoire of boundary work across fields consisted of cross-field referencing, building up ties, arena switching, circumventing of formal procedures and invading other fields. Challengers in the domestic field referenced changes in the international regulation of agency work as a justification for change in the domestic field. In Germany, the challengers cited the Netherlands to indicate how agency work could be beneficial for labor market performance during the periods of building up pressure and power reversal (Bundesregierung, 2002; IAB, 1997; see also van Gestel & Hillebrand, 2011). Building up ties is observed when challengers established their own international federation(s). The temporary agency industry formed its international federation of private employment agencies (Ciett) in 1967 with affiliates in various countries (Ciett, 2013a). Arena switching is observed when challengers moved with their ambitions to related fields, where they expected less opposition. In the case of temporary work, this is exemplified by agencies’ coordinated lobbying in the transnational field, which finally led to a breakthrough in getting the ILO convention 181 (details below). Another variety was bypassing formal procedures by using alternative platforms and forums at the transnational level to launch challenges to single countries and international organizations (Blanpain, 1999; Ciett, 2000). Invading other fields was carried out by multinationals when they set up subsidiaries in countries at a time when national regulation had not yet been changed. As early as 1965, Manpower established its first subsidiary in Germany, well before the change in legislation allowing agency work in 1972.

Changing a field’s settlement by challengers redrawing boundaries

The most important condition for a challenge to arise is the defeated parties staying in the game in the first place. By enacting the “fragments and elements of alternative systems that represent the legacies of constitutional struggles over order and social movements whose settlements or defeats helped fix the path that triumphed” (Schneiberg, 2007, p. 48), the mere survival of defeated actors already alters a field, as it shows (also to other challengers) that there are niches that are not completely under the control of the incumbent coalition. From there, former incumbents contribute to move a stable field settlement into periods of building up pressure and power reversals when they participate in the organizing of a challenge as they appropriate and enact their repertoire of the politics of boundary work. Certainly, there are additional context conditions contributing to a change in a field’s settlement, for example, economic ideologies, waves of restructuring or substantial problems such as high unemployment (Doellgast & Greer, 2007; Palier & Thelen, 2010). However, these context conditions only become effective inasmuch as (new) challengers enact them by means of successful boundary work like appropriating organizational vehicles or building up ties with the transnational field.

In the field of agency work, during the post-war period and during the building up of pressure, work agencies had to stay in the game in a rather hostile regulatory environment. Agencies, for their part, also had to appropriate organizational vehicles by setting up their own associations apart from existing business associations; they had to legitimize their challenge symbolically and lobby the legislator. In this, the agency industry primarily moved from presenting agency work as an instrument of flexibilization to depicting it as a solution to integrate excluded groups without negative effects for other groups in the labor market. In parallel, the agency industry also blurred the symbolic boundaries around its business model, making it ‘opaque’ (Briscoe & Murphy, 2012) and difficult to attack as its appearance oscillated virtually between public-private employment agencies, private job placement, agency work and mere subcontracting. From these considerations I conclude:

Changing a field’s settlement by challengers crossing boundaries

In the long-term struggle over legalizing agency work in Germany, the most important turning points were regularly preceded by successful boundary work across fields by challengers. I found numerous examples of how the agency industry engaged in crossing boundaries to the transnational field (Ciett, 2000, 2012; ILO, 1997c): linking up to the transnational field allowed the infant agency industry to exploit connections to successful pioneers in other domestic fields, above all, to those in the United States (Wenger & Kalleberg, 2006). As Hatton (2011) argues, the seeds of change were planted in the US by a new generation of agencies founded in the late 1940s. These agencies developed early templates for how to justify the business model in a hostile environment, as in the 1940s many US states had strict regulations on agency work (Freeman & Gonos, 2005; Hatton, 2011). Today, one exemplar of this generation, Manpower (founded in 1948), is established among the world’s largest agencies (see Table 2).

OPEN IN VIEWER

Table 2.

World leaders in the PEA industry, 2012.

Name	Country of origin (founding year)	Agency workers, placed on assignment per day on average, 2012	Permanent staff, 2012	International locations, 2012	Market entry in Germany
Adecco	Switzerland (Adia, 1957), France (Ecco, 1964)	700,000	32,000	63	1962
Manpower	USA, 1948	636,600	26,000	80	1965
Randstad	Netherlands, 1960	581,700	29,320	39	1968

Source: Companies’ annual reports 2012.

Agencies’ crossing of the boundaries between the domestic and the transnational fields increased the pressure on the post-war settlement from the late 1960s on: the foundation of Ciett in 1967 was triggered by the Swedish government’s request to the ILO to uphold its legacy on private work agencies that year (Vosko, 1997). Before that episode, the agency industry had already established several important bridgeheads in Europe, in countries such as the Netherlands (where Randstad was founded in 1960) and Switzerland (where one of the predecessors of Adecco, i.e. Adia, was founded in 1957), and also in France, the UK and Belgium (Ahlberg et al., 2008). In the German field, the most indicative episode was the Federal Court decision of 1967, in which the German subsidiary of the Swiss company Adia (established in 1962) won a Federal Court decision, which led to the Temporary Agency Work Act (1972). Thereby, an intransigent incumbent coalition was forced to adapt to the challengers’ initiative launched from the outside.

Having already established cross-border ties through informal platforms and debating forums, the agency industry was well prepared to ‘orchestrate’ (Ciett, 2000) its second major cross-border challenge between the early 1990s and 1997 by switching within the transnational arena between the EU and the ILO, first challenging the dominant interpretation of the European law (ECJ, 1991, 1997) and then dominating the policy process around the ILO convention 181 (ILO, 1997c). In this situation, the representatives of the incumbent coalition from Germany had a difficult time making their voice heard in the international regulatory debate (Blanpain, 1999), which had already shifted from the unions’ concerns for workers’ protection to employers’ views favoring labor market flexibility in order to restore manufacturing firms’ competitiveness. In addition, the unions’ repertoire of acting across fields was much less developed and relied on the formal procedures of international labor diplomacy at the ILO and within the EU (Ahlberg et al., 2008; EuroFound, 2006; Håkansson et al., 2009; Vosko, 2010). Those representing the (still) challenging group within Germany could stage their views prominently, like, for example, the HR director of Volkswagen and later head of the reform commission, who had already been a member of the advisory board of Ciett’s project for promoting the industry (Ciett 2000). From this I conclude:

Change in a field’s settlement through tensions in incumbents’ internal boundary work

From an incumbent’s perspective, there is a trade-off between activating and reinforcing internal boundaries and the upholding of a field’s boundaries by accommodation and compromise. Whereas in the consolidation of the post-war settlement (1949–66) incumbents predominantly engaged in intensifying field governance up to the point of excluding agency work altogether, they could only stabilize the field’s settlement by accommodating the agencies’ challenge (e.g. the Temporary Agency Work Act, 1972) in the period of building up pressure. Each compromise, however, had repercussions in that it moved the field’s original settlement gradually towards further compromises, exceptions and deregulation (Sengenberger, 1984; Vitols, 2004). Combined with the erosion of the post-war coalition, these exceptions allowed the agency industry to extend its market niche slowly (Bellmann, 2004). In that process, incumbents’ practices began to contradict each other. For example, parallel to increasing unemployment rates in the 1990s and the Federal Court decision of 1994 allowing private job placements, the agency industry invented a public-private partnership model that signaled to unions that agency work would function as a complement to public labor market policy, and that the unions would be included in managing this instrument (DGB, 1996a). In parallel, negative attitudes towards agency work were softening among unions as the agencies signaled a policy to avoid replacing strikers (Bronstein, 1991). In the end, the unions were open to a compromise in order to sustain their overall position within the field. The German unions changed their long-held demand for abolishing agency work in principle (DGB, 1996a, 1996b, 1996c; Wölfle, 2008) and suggested facilitating the legalization of agency work (DGB, 1996a); mere ideological reinforcement of an industry ban (as in DGB, 1981) became contradictory to a compromising practice:

Change in a field’s settlement through tensions between upholding and protecting boundaries

Protecting external boundaries can also become contradictory to upholding a field’s settlement from within; for example, when incumbents reach out to the transnational field for justification (as with the post-war incumbents’ upholding of the ILO legacy) and are then confronted with the necessity to accommodate challengers who can refer to changes in the transnational field. This holds in particular if the field of reference’s settlement is weakly enforced, as is exemplified by the ILO, which is intellectually and normatively influential, but lacks authority in enforcing its conventions (Rodgers et al., 2009). For example, the early conventions banning agency work were only ratified by 11 (Conv. 34) and 42 (Conv. 96) member states, respectively (ILO, 2009; Vosko, 1997). Given an already weakened incumbent coalition, the German unions’ decision to abandon the demand to prohibit agency work – according to a labor market policy officer involved in representing the DGB in the debates around the ILO convention 181 – anticipated a power reversal in the transnational field, not least because of the European Court decisions liberalizing private employment agencies (ECJ, 1991, 1997). However, as documented in the ILO protocols of that session, initially the German union representatives were among those resisting the new convention (ILO, 1997c). This episode reveals how the upholding of the field’s boundaries in the German field ran into contradiction with a weak performance of incumbents to protect the external boundaries within the transnational arena:

The politics of boundary work and non-extinctive change in organizational fields

According to the preceding analysis, incumbent coalitions dominating even the most stable of fields cannot preserve a given settlement for ever. The rules working in favor of the incumbents at one point in time might culminate in their reversal when former challengers take precedence, and vice versa. One important source for such a power reversal is that collective actors might lose one battle but are rarely defeated entirely, so they have the chance to learn from their defeats and find ways – including via the transnational field – to return as successful challengers in later periods (Thelen, 2009). Likewise, the politics of boundary work illuminate how even incumbent coalitions that appear powerful and encompassing in the short and medium term are not capable of preserving a settlement in the long run, because they lose control over the definition of the combat arena, in particular when they are not capable of protecting the field’s external boundaries.

In other words, the politics of boundary work are likely to be more important for institutional dynamics when the process under study is non-extinctive in nature, as the conflicting institutional logics are so deeply embedded that conflict builds up over and over again, albeit with changing actor constellations. This holds true in particular for those organizational fields that group around institutions that decommodify the fictitious commodities of labor, money and natural resources (Kalleberg, 2009; Polanyi, 2001; Streeck, 2009), as institutional dynamics around these – like geophysical plate tectonics – unfold over rather long time periods (Kieser, 1994; Thelen, 2009). However, societal as opposed to geophysical plate tectonics bear the potential to shake up a field’s settlement only as far as the defeated parties are capable of connecting to past legacies and other fields.

It is fair to state that the repeated game situation of non-extinctive institutional change is not well understood in organization theory, because studies focus overwhelmingly on field dynamics in an internally driven equilibrium framework of continuous institutionalization (for exceptions, see Davis & McAdam, 2000; Kieser, 1994; Meyer et al., 2005; Schneiberg & Soule, 2005). This also points to the limitations of this paper. As an analysis of one particular field, the preceding discussion cannot establish the conditions and mechanisms of non-extinctive institutional dynamics completely. For example, what are the identifying properties of non-extinctive institutional dynamics apart from mere survival of defeated collective actors? And how can non-extinctive institutional dynamics be separated from other dynamics like institutional entropy or cycles of continuous institutionalization? Likewise, it remains unclear whether boundary work is played out differently in fields that follow other patterns of institutional change. A few important expectations can be expressed, however. Above all, non-extinctive institutional dynamics are expected to occur when there is a long-term mutual interdependence of a field’s actors and deeply rooted political division between opposing societal logics touching upon social identity and material interest at the same time.

Despite the shortcomings, however, it is suggested that the politics of boundary work can be used as one approach to tackling non-extinctive institutional change by examining more closely when and how institutional dynamics are modulated by continuous controversy enduring over long periods rather than being caused by short-term upheavals (cf. Fligstein & McAdam, 2012). Although the events after the 2004 legalization of agency work require an analysis of their own, they also point in that direction as agency work has remained a controversial issue in Germany (Promberger, 2012). The “new kings”, i.e. the coalition between the agency industry, business leaders, political parties and their allies in the unions, taking over the field in the early 2000s, came under pressure very soon. Not only did the new incumbents have to face immediate mass protests on the streets (“Montagsdemos”, Indymedia, 2004), the pressure also spilled over into the field of electoral politics, where the power constellations changed in the federal elections of 2005 (Dressler, 2012; Müntefering, 2012). In public campaigns, the unions frequently bring to attention the unequal treatment in major client companies, severe rule violations by single agencies, and loopholes in the law (IG Metall, 2012; ver.di, 2011a); as one reaction, a collective minimum wage contract was put into effect by the federal government (BDA, 2010). Also, the unions realized that agency workers can be organized around client companies, i.e. crossing the boundary to the neighboring field (Benassi & Dorigatti, 2014). In 2011 the unions won a change of law that restricted in-house contracting, i.e. large corporations replacing regular workers by engaging agency workers via their own subsidiaries (ver.di, 2011b); and in 2013, in a couple of client industries’ collective agreements, agency workers’ standards were raised through industry-specific wage add-ons (BAP & IGZ, 2013; IG Metall, 2012). Last but not least, the new challengers cross the field’s boundaries by referencing extreme exploitation of agency workers abroad and by rebuilding transnational ties in the issue field of contract labor (Ciett, 2013b; ILO, 2006;). However, still being very new, the transnational (re)positioning of organized labor is hampered in the field of agency work. This is due to a disagreement between those unions trying to establish social dialogue with the industry, i.e. mostly around UNI, the global union federation for private services (Ciett Corporate Members Committee & UNI, 2008; ETUC, 2009; EuroCiett & Uni-Europa, 2007; UNI, 2009), and those unions that still want to reduce the scope of the agency industry as far as possible, i.e. union federations representing workers in the client industries such as IndustriAll, the global federation of manufacturing unions (CGU, 2010; IndustriAll, 2012).

The (micro-)politics of boundary work and the categorization of work(ers)

As field boundaries constitute organizations’ social environment, work on field-level boundaries is linked to the (micro-)politics shaping organizational boundaries. Hence, a study on legalizing agency work at the field level also informs about the construction of distinctions between groups of workers within organizations (Lawrence, Leca, & Zilber, 2013) through the micro-politics of boundary work. Within organizations, categorizing work is enacted in encounters between individuals belonging to different social groups with repercussions for work organization and distribution among these groups (Håkansson & Isidorsson, 2012). Being an agency worker means, first and foremost, being placed in an outsider’s role not easy to overcome, given permanent coworkers’ need for distinction (George, 2003; Purcell, Brook, & Lucas, 2011). As Gray and Kish-Gephart (2013) have argued, in organizations various practices perpetuate these work-related boundaries, thereby imposing a stratification of organizational members into roles and status positions. These categorizations, then, might even appear as objectified organizational practices enshrined in the “routines, policies, and practices that organizations employ to recruit, evaluate, hire, and retain employees” (Stainback, Tomaskovic-Devey, & Skaggs, 2010, p. 230) and legitimized through symbols, rituals and myths obscuring the boundaries’ ideological origin (Bendix, 2001 [1956]). Even more, organizational incumbents, most frequently management, might reach out to the regulatory field to bring about regulatory change in support of their categorizations of workers.

Based on the preceding analysis, it is suggested to include “collective class work” (Gray & Kish-Gephart, 2013, p. 675) from a worker’s perspective more explicitly into the theoretical debate on how boundaries around work are drawn. With respect to categorizing work from a worker’s perspective, collective action becomes a true game changer, in that organized workers can not only issue counter-narratives from below, but also resist management (Spicer & Böhm, 2007) by challenging the employers’ “divide and conquer” strategizing (Kieser, 1994, p. 614). Thereby, workers appropriate the organizational vehicles necessary to alter organizations’ practices and to install the institutionalized safeguards of labor law and industrial relations, which force both actor groups, i.e. employers and organized labor, to negotiate over the categorization of work (e.g. Helfen & Sydow, 2013).

In this sense, it is also fruitful to contrast the legalizing of agency work with studies on the de-institutionalization of permanent employment (Bidwell, 2013; Ahmadjian & Robinson, 2001). The legalizing of agency work certainly intervenes in this de-institutionalization by weakening institutional safeguards supportive of the long-term employment relationship and giving employers a “weapon” to avoid unions (DGB, 2009; Hatton, 2014). And where institutionalized labor relations erode through the introduction of (new) organizational practices to hire labor – as has been expected for Germany (Palier & Thelen, 2010; Streeck, 2009) – future struggles might resemble the labor disputes of an earlier, pre-institutionalization era. However, as long as institutionalized safeguards are not eradicated completely, they might also be deployed by workers to mobilize collectively in order to challenge work categorizations (Heery, 2009; Wills, 2009).