Extending Meta-Organization Theory: A resource-flow perspective

Meta-organization theory has shied away from a systematic consideration of the complexities and consequences of resource acquisition patterns, instead assuming member organizations furnish the resources their associations require. The theory reflects empirical enquiry, which has focused primarily on the most visible form of resources, the association’s finance budget and staffing. This paper develops a conceptual framework to show the wider range of ways in which meta-organizations acquire resources and presents a resource-flow perspective. We utilize insights from resource dependency theory to specify two dimensions of resource flow. The first addresses the source of resources, distinguishing those acquired from the meta-organization’s membership from those acquired from external interests. The second concerns whether the meta-organization secures control over the use and allocation of acquired resources, or not. We identify four resource flows utilizing these two dimensions: member resourcing, associational resourcing, contributed resourcing and generated resourcing, and discuss how each resource flow relates to meta-organizational activity, highlighting when a particular flow can be expected. The resource-flow perspective allows us to adapt existing typologies to define a conceptual space onto which variations among meta-organizations can be mapped. This space focuses on the expected level of resource engagement of members in the meta-organization and the extent of resource contribution from interests in the meta-organization’s environment. Finally, we discuss areas for development of the resource-flow perspective and its potential to support future research.


Introduction
Meta-organization theory has developed since it was initiated by Ahrne andBrunsson (2005, 2008) to complement organization theory and its preoccupation with individual-based, hierarchical organizations.Ahrne and Brunsson (2008) argued that, like all formal organizations, meta-organizations are decided social orders.However, they have a distinctive character, being associational in form and with a membership composed of organizations.The main focus of theory development has been on validating and elaborating the defining features of a meta-organization.Although recognized as an important issue to explore, less attention has been paid to the ways in which metaorganizations vary (Ahrne & Brunsson, 2008;Berkowitz & Bor, 2018;Berkowitz, Brunsson, Grothe-Hammer, Sundberg, & Valiorgue, 2022).Garaudel (2020) proposed two approaches to theorizing such variation.One approach is to focus on meta-organizations according to their institutional niche -local associations of voluntary organizations, national business associations, or international trade union associations, for example.Each will have a specific configuration of characteristics that will be significant in understanding why they have emerged, how they operate and how they are sustained.A second approach is to identify generic dimensions of meta-organizations that are theoretically robust and reveal similarities and differences between them across their institutional settings.The current research adopts the latter approach.
One dimension of variation identified is resourcing (Berkowitz & Bor, 2018;Bor, 2014;Bor & Cropper, 2016).While acknowledging that resources are intrinsic to an organization, Ahrne and Brunsson (2008) assume that, in the case of meta-organizations, the availability of resources is essentially unproblematic and therefore leave that area largely unexplored.However, studies of meta-organizations, including those by Bor (2014), Greenwood (2002) and Karlberg (2019) suggest that the acquisition of resources can be more complex and of greater consequence than metaorganization theory suggests.
The current research seeks to conceptualize patterns of resource acquisition in meta-organizations and asks whether differences in resource acquisition might contribute to differences among meta-organizations.We utilize resource dependency theory (RDT) (Pfeffer & Salancik, 1978) as it explores the process of the acquisition of resources to secure the survival of organizations.However, RDT is based on the needs of an individual-based organization rather than an association of organizations.This article applies aspects of RDT selectively to develop the concept of resource flow, which we define as a way in which resources are secured by the meta-organization to enable its continuing presence and activity.The perspective admits the full range of resources needed by organizations and we follow RDT's central insight to include resource flows between the metaorganization and the broader environment with which it interacts.Expectations about patterns of resource engagement with the meta-organization by members and external sources, respectively, will differ, not least as a consequence of the specific constitution of the meta-organization and the privileged relationship with its membership (Bor, 2014).The distinction made in RDT between securing access to resources and having control or discretion over resource allocation and use is important here.Pfeffer and Salancik (1978) argue that participants in any organization will separate their commitments to the organization from commitments elsewhere.Consequently, it is important to distinguish resources that member organizations are expected to provide to the metaorganization -as part of what we refer to as the membership dealfrom those sought, offered, made available, or acquired beyond that deal.The paper and concept of membership deal also draw on the argument in RDT that activity (for which Pfeffer & Salancik, 1978, p. 86 use the terms event and outcome) affects resource flow.Where meta-organization theory defines the boundary of the organization as its membership, RDT defines the organization in terms of its activities and behaviours and conceptualizes its boundary as the point at which activity is no longer controlled by the organization, but by other interests.In this way, RDT emphasizes the analysis of the pattern of interests in the organization (see also Wry, Cobb, & Aldrich, 2013).In the meta-organization context, these include the interests of members, with their privileged status, and also the interests of actors in the broader environment of relevance to the meta-organization.This conceptual paper highlights differences in the nature, significance, and scope of meta-organizational activity and the influence of these factors on the development (and choice) of resource flows.
The perspective also allows us to adapt existing typologies of organizations of organizations (Ahrne & Brunsson, 2008;Provan, 1983;Provan & Kenis, 2008), and the resulting framework specifies how meta-organizations can differ in resource-flow terms.We discuss some key implications of the perspective for meta-organization theory: specifically, we reconsider the conception of boundary, activity and member participation.Following from this, we indicate how the resource-flow perspective could provide a foundation for future research into these aspects of meta-organization.Ahrne and Brunsson (2008) defined meta-organization in terms of three essential characteristics.The first is that meta-organizations are decided social orders, highlighting the difference between what first Blau and Scott (1963) and later Ahrne (1994) called formal and social organization.Formal organization stems from decisions on issues such as boundaries, goals and rules.In contrast, social organization originates with institutions and networks and is an emergent social order present in all organizations.Ahrne and Brunsson (2008) emphasize the formal character of a metaorganization and note that they are like other formal organizations in being attempts to display collective rationality.From inception, a meta-organization represents a distinct and additional organization, and, like other organizations, meta-organizations have 'a certain degree of autonomy. ..they can own and control some resources.If an entity lacks autonomy or cannot control its resources. ..it would not be seen as an organization in its own right' (Ahrne & Brunsson, 2008, p. 47).

Meta-Organization Theory and Resource-Related Assumptions
The second defining characteristic of a meta-organization is that its members are organizations rather than individuals.When developing meta-organization theory, Ahrne and Brunsson (2008) focused particularly on explaining the implications of this difference.One implication is that organizations generally have access to far more resources than individuals, which led to an assumption that remains current in meta-organization theory: 'members of a meta-organization can supply it with a great capacity for action and access to resources right from the start' (Ahrne & Brunsson, 2008, p. 142).It follows that meta-organizations may 'become strongly dependent on their members in order to obtain resources' (Ahrne & Brunsson, 2008, p. 143).Finally, Ahrne and Brunsson (2008) point out that the 'power and resource access differences among organizations are considerably larger than the corresponding differences among individuals' (p.144), which has consequences for the third characteristic.
The third essential characteristic is that meta-organizations are associations.The term relates to both the relationships between members of the meta-organization and those between members and the meta-organization.The relationship between members is essentially horizontal (Warren, 2001) or symmetrical (Grandori & Soda, 1995), such that members of the meta-organization are expected formally to have equal membership rights and obligations: 'All members are expected to be of equal value and, in principle, to exert equal influence' (Ahrne & Brunsson, 2008, p. 11).In practice, those expectations may be moderated by other factors, including patterns of resource flow.The relationship between members and the meta-organization has two important facets.Membership of the meta-organization is voluntary, which weakens the meta-organization's authority (Ahrne & Brunsson, 2008).A meta-organization is also member-governed, which is viewed here from a cooperation perspective (Gulati, Wohlgezogen, & Zhelyazkov, 2012).Governance from a cooperation perspective concerns decisions about the membership boundary, goals and control of direction and behaviour (Bor, 2014), including the contributions to be made by members and the outcomes they can expect in return.Governance through a cooperation lens is aimed at stability and goal attainment and will also include decisions about resource dependencies and investment risks (Gulati, Wohlgezogen, et al., 2012).Provan and Kenis (2008), using the term member-governed in their discussion of whole networks, adopt a different perspective, a coordination perspective in Gulati, Wohlgezogen, et al. (2012) terms.A coordination perspective focuses on the management of task interdependencies and uncertainties, including planning, organizing, motivating, coordinating and monitoring (Bor, 2014), aiming at efficiency, effectiveness and flexibility or adaptiveness through joint action (Gulati, Wohlgezogen, et al., 2012).Ahrne and Brunsson (2008) use the term management to cover both perspectives and all tasks.For clarity, we adopt Gulati, Wohlgezogen, et al.'s (2012) distinction, using the term governance in relation to a cooperation perspective and management to reflect a coordination perspective.Resources are important considerations in each.
The associational nature of the meta-organization led Ahrne and Brunsson (2008) to argue that members are likely to find a shared interest in its purpose.They recognize, however, that some members may join only to maintain a watching brief, with limited commitment to the organization's purpose and its activity; nevertheless, they would still be expected to provide the resource concomitant with membership.Free-riding, they suggest, is unlikely to be a problem, owing to the low costs and opportunities to exert influence.Nevertheless, the capacity to meet the terms of the membership deal can still be problematic.Accordingly, 'dissimilarities in access to resources can mean that some members find it difficult to play a full part in meta-organization activities' (Ahrne & Brunsson, 2008, p. 145).Such differences can raise questions over the relative status of members.Ahrne and Brunsson emphasize that deliberations over the appropriate forms of decisionmaking and voting rights are likely; with larger contributors seeking a greater say in decisions.Equally, meta-organizations can promote solidarity, as in international labour organizations, where larger members with greater resources support the activities of the association when other members cannot (Garaudel, 2020).
Meta-organization theory is neither the first, nor the only attempt to theorize inter-organizational arrangements.While attempts have been made to clarify differences among inter-organizational arrangements (see, for example, Alter & Hage, 1993;Grandori & Soda, 1995;Provan, 1983), relatively few of the concepts developed have been used in continuing theory development.The concept of an inter-organizational network has been the exception (Ebers, 1997;Kenis & Oerlemans, 2008).Although the extent to which inter-organizational networks are organized and thus classifiable as meta-organizations is debated (Ahrne & Brunsson, 2008, 2019), a distinction has been recognized, separating serendipitous networks (emergent social orders) and goal-directed networks (decided social orders) (Kilduff & Tsai, 2003).Goal-directed networks have assumed prominence particularly in the public management literature, labelled as whole network, multiorganizational network, or multilateral network (see e.g.Auschra & Sydow, 2022;Human & Provan, 2000;Provan & Kenis, 2008;Provan, Kenis, & Human, 2008;Saz-Carranza & Ospina, 2011).These forms of organized network are perhaps the most proximate to that of meta-organization.Like meta-organization, the goal-directed, multi-organizational network is defined as a decided social order and the members are organizations.However, the concept does not differentiate between symmetric and asymmetric inter-organizational arrangements.Grandori and Soda (1995) suggest that this distinction is important for theory development, and meta-organization theory starts from the principle of symmetric inter-organizational arrangements, specifying the relation of members to the collective as one of association (Ahrne & Brunsson, 2008).Furthermore, as Berkowitz et al. (2022) highlight, Ahrne and Brunsson's concept of meta-organization should not be read as equating to that discussed by Gulati, Puranam and Tushman (2012).The latter is a near synonym for concepts such as networks or ecosystems, populated both by individuals and organizations and including both emergent and established social orders.It is thus significantly different and largely incompatible with Ahrne and Brunsson's conception of association expressed as formal organization.Even when they address decided order, Gulati, Puranam, et al.'s (2012) version of meta-organization is one formed and governed by a single organization to serve its own interests (e.g.internet exchange platforms such as Amazon) and does not focus on the association of organizations.
The precise specification of the properties of meta-organizations has helped theorization.Although meta-organizational studies may develop insights relevant to other inter-organizational forms and vice versa, careful consideration of the effect of their constitutive differences is warranted.Having clarified the similarities and differences with other organizational arrangements, we turn to our main concern, the development of a resource-flow perspective.

Resource Theories and the Concept of Resource Flows
Meta-organization theory has paid little systematic attention to resources.While membership fees are well-recognized (see e.g.Ahrne & Brunsson, 2005;Berkowitz & Dumez, 2016;König, Schulte, & Enders, 2012), other types of resource are rarely mentioned, and may be considered more as a cost for members than as a resource for the meta-organization (see Dumez & Renou, 2020 for example).Discussions of inter-organizational collaboration have taken a more elaborate view.Resources are commonly distinguished by their type, though without specifying, theoretically or analytically, the basis for the differentiation: financial resources are separated from human resources and those in turn from material resources and from symbolic resources (e.g.Bardach, 1998).Resource listings can be extended to include fine distinctions between types of resource or be brought together under a few essential headings, but, as Barney (2001) argues, the search for a definitive listing is not ultimately helpful.First, those resources and resource mixes that have value will depend on the specific conditions in which a meta-organization emerges and develops, and the nature of its work.Second, a focus on the analytical properties of resources rather than conventional distinctions made in practice may be more useful for theory development.For example, the distinction between tangible and intangible resources has been closely examined in whole network studies (see Huang & Provan, 2007;Provan & Huang, 2012).
The resource-based view (RBV) (Barney, 1991(Barney, , 2001) ) and its offshoots, including the theory of dynamic capabilities (Ambrosini, Bowman, & Collier, 2009;Eisenhardt & Martin, 2000;Schilke & Goerzen, 2010) and recent discussions of resourcing as organizational processes drawing on practice theory (Auschra & Sydow, 2022;Deken, Berends, Gemser, & Lauche, 2018;Feldman, 2004;Feldman & Worline, 2012;Schneider, Bullinger, & Brandl, 2021), depict important links between resources and the character, capability and effective functioning of organizations.These theories emphasize the creative and effective use of available resources within the organization to meet the organization's goals or, in RBV terms, to produce value.In terms of resource usage, extant research suggests that the important properties are those of particularity and stickiness, transferability, fungibility, substitutability and divisibility, and whether the resource can be readily transferred, combined and aggregated (Barney, 1991(Barney, , 2001;;Foa & Foa, 2012).There is considerable potential to use these theories to understand how the translation of resources into meta-organizational capability is achieved dynamically, either routinely or strategically.
The focus of this paper is, however, different.It examines the ways in which necessary resources are acquired by, or made available to, the meta-organization.This is the process we conceptualize as resource flows that result in particular patterns of resourcing within the meta-organization.We draw on the arguments of RDT (Pfeffer & Salancik, 1978) to accomplish two aims.
First, Pfeffer and Salancik's (1978) focus is on the external sourcing of resources by organizations, highlighting that any organization depends on its environment to acquire its resources.The task is to limit the constraints arising from resource dependencies.In their account of meta-organization, Ahrne and Brunsson (2008) focus on exploring one type of activity -boundary work focused on members' identity and status.The argument is that meta-organizations can establish a presence and sustain activities that meet their members' interests in a self-contained and self-sufficient way.The resources they need will be available from members.However, it is recognized that meta-organizations also engage with organizations beyond their membership to garner resources.Two of the motivations for establishing a meta-organization that are widely discussed involve exchanges between the meta-organization and external (non-member) interests.One rationale is that members see collective voice and action as a means of influencing their shared environment.Ahrne and Brunsson (2008) present a second rationale as something of a special case: powerful interests in the environment might encourage the formation of a meta-organization.Either type of exchange might involve resource acquisition in which external interests provide material or symbolic resources to the meta-organization.Accordingly, we draw on RDT to argue that meta-organizations will also be expected to seek, receive and access resources from their environment and to reciprocate in relations of exchange.
While the meta-organization and actors in its environment may indeed have certain resource connections, Ahrne and Brunsson (2008) also point to the privileged relationship that exists between the meta-organization and its members.Certainly, the meta-organization is its membership.Nevertheless, as Pfeffer and Salancik (1978) argue, an organization will only command a part of its members' attention and commitment.When a meta-organization is formed, a critical decision, therefore, concerns the level of resource engagement with the meta-organization expected of member organizations.That specification of resource engagement forms part of the membership deal, which, inter alia, establishes a shared expectation of the scope of purpose, activity and ambition of the meta-organization, secures resourcing for essential functions and activities, and indicates the benefits members can expect in return.The membership deal is often set out explicitly in contracts, activity and resource plans, and collective decisions, but it can also develop through custom and practice and hence may be implicit and contested.The latter represents what Pfeffer and Salancik (1978, p. 72 ff.) refer to as the enacted environment -referencing the experience of the past to make sense of what is expected now and in future.This is also a measure of a social actor's agency, in Emirbayer and Mische's (1998) terms.Maintaining the membership deal is thus an ongoing governance responsibility of the member organizations that asserts the significance of the meta-organization to members.
Meta-organizations can also expect to receive and acquire resources that are not specified in the membership deal.While member participation in the meta-organization can be viewed as a duty and a defined resource contribution (e.g.attending meetings of the governing body), it can also be viewed as an entitlement, which may or may not be taken up.When viewed as an entitlement, participation may represent an additional, discretionary input into the meta-organization, for example, to lobby for activities of interest to the member, or (more generously) to underwrite the metaorganization and its activities when others are unwilling or unable to contribute.
Non-members whose interests are served by the meta-organization may also provide resources.Equally, the meta-organization may express shared interests that can only be pursued through engagement with actors in the environment, or with members to secure additional resources outside the membership deal.The resource flows that fuel such relationships can include the full range of resources that organizations require to thrive and survive -financial support, expertise, access to power centres and symbolic resources.Those agreements can be extended and amended over time, but these flows are likely to be less predictable and less stable over time than those defined in the membership deal.More importantly, perhaps, the terms of each contribution may be subject to negotiation in a way that the terms of the membership deal are not.Yet, contributions may also be made without the knowledge of the meta-organization; for example, where members or external organizations endorse the meta-organization's work (see e.g.Stuart, Hoang, & Hybels, 1999).
Second, RDT asserts that organizations may gain access to resources without necessarily obtaining full discretion over their use or allocation: '. ..discretion is rarely absolute.More commonly, there are degrees of shared discretion' (Pfeffer & Salancik, 1978, p. 48).Decisions and other consequential activities of the meta-organization may, then, be subject to the conditional commitment of resources by members.Commitments made by representatives at the meta-organization's governing body, for example, may be subject to availability, or subject to use in particular ways when it comes to the practicalities of coordination.Indeed, Ahrne and Brunsson (2008) observe: The meta-organization is dependent for many decisions not only on its members' passive consent but also on their active involvement, which means each member must engage its own members. . .The decisions may also demand that members are willing to cooperate and to contribute with their own resources.(pp. 114-115) Resources the meta-organization may not fully control would include in-kind, non-monetary contributions such as staff time or meeting spaces.Meta-organization theory tends to define such input as a membership cost (e.g.Dumez & Renou, 2020) instead of considering them as resources available to the meta-organization.However, these resources can come with disadvantages for the meta-organization.Where member organizations provide staff time for the meta-organization on a goodwill basis, they will ultimately reserve the right to decide who will be made available, when, and for which activity.Uncertain resource availability is likely to have important consequences for the coordination of collective efforts (Grandori, 1997) at both management and governance levels.
In sum, the acquisition of resources by the meta-organization has two important dimensions: the source of resources made available to the meta-organization; and the location of control over the allocation and use of those resources.Table 1 shows that four types of resource flow can be derived from these two aspects.Each resource flow is discussed in more detail below.

Member resourcing
Member resourcing refers to a resource flow in which member organizations make resources available in support of the collective effort.In doing so, they fulfil expectations of resource engagement set out in the membership deal.However, each member retains control over the allocation and use of the resource it provides.Member resourcing often takes the form of contributions in-kind, for example, a representative nominated to serve on the meta-organization's governing body, staff time to support committee work, participation in activities, meeting spaces, and access to other services (e.g.Brouder, 2009c;Cropper & Bor, 2018;Nowrot, 2009b;Porter, 2009;Varney, 2009).The acquisition of member resourcing to enable the coordination and realization of activity is particularly sensitive, since members may see this as incurring substantial costs on behalf of the collective.Meta-organizations may therefore apportion the resources required among members, or responsibility might be rotated among members over time, or some combination of the two.For example, the agency that holds the chair of the Consultative Committee for Space Data Systems (Pilgram, 2009) provides all secretarial, administrative, technical and infrastructural support required during its year in office; however, the activities involve all (or a subset of) members.Similarly, the country presiding over the G8 covers all resources needed for the preparation and organization of meetings during its presidency.Brouder (2009b) points out that the approach can lead to considerable variety in the resources made available.

Associational resourcing
Associational resourcing refers to a resource flow into which member organizations release resources to the meta-organization to meet the terms of the membership deal.In this case, they also transfer discretion over the allocation and use of those resources to the collective, the meta-organization, so that the resources effectively form a pool.While various types of in-kind resources can be acquired from member organizations, such as personnel seconded to the meta-organization (see Quick & Brouder, 2009), the most common type of resource transferred through associational resourcing is an annual membership fee.Financial contributions are easily transferred and reckonable; they accumulate and aggregate, and are flexible to use, making resourcing specifiable and easily monitored.In-kind contributions can be more difficult to transfer and equate, and are usually

Member resourcing
Members make resources available as part of the membership deal with the meta-organization but retain control over their allocation and use

Association resourcing
Members allocate resources to the meta-organization as part of the membership deal and pass control over those resources to the metaorganization External, outside membership deal

Contributed resourcing
Parties contribute resources outside the membership deal but retain control over their allocation and use

Generated resourcing
Parties provide resources outside the membership deal and pass control of those resources to the metaorganization transferred for an agreed period (Rothschild-Whitt, 1976).They may also require additional agreements to govern the transfer of control, e.g.secondments of personnel.

Contributed resourcing
Contributed resourcing refers to a resource flow in which the meta-organization acquires access to resources from beyond its membership deal and where the contributor retains the discretion to determine the availability and terms of use of the resources it contributes.This resource flow often remains hidden, and, as Baker and Mathews (2009, pp. 379-380) note, 'the reliance on voluntary contributions represents a weakness in the independence' of the meta-organization.Like member resourcing, contributed resourcing is most often in-kind -a meeting space (Porter, 2009), expertise (Hallström & Higgins, 2009), or administrative, financial or audit services (Wälde, 2009).Other important examples of contributed resourcing include legitimacy, esteem and trust (Berkowitz & Souchaud, 2019;Greenwood, 2002).

Generated resourcing
Generated resourcing refers to a resource flow through which the meta-organization acquires resources from outside its membership deal and the right to decide on the allocation and use of those resources.As with associational resourcing, such resources are not necessarily provided unconditionally, but can be expected to be subject to the terms of the agreement with the resource provider.The meta-organization can be made accountable for the use of resources or for delivering promised outcomes via an exchange contract (e.g. for services, see Nowrot, 2009a;Schreuer, 2009), grant agreement (see e.g.Bor, 2014;Nowrot, 2009b), or other mechanism such as scheduled reporting (see e.g.Karlberg, 2019).Nevertheless, donations or gifts can be offered (e.g.Brouder, 2009a;Koh, 2009) to support particular projects, or the organization in general, that have few strings attached.As with associational resourcing, the resource provided through generated resourcing is often financial.

Activity and Resource-Flow Interaction
Meta-organizations are efforts to govern, making decisions, providing directions and mandates, and sometimes issuing resolutions on sanctions.However, they also exist to generate, orchestrate and maintain a programme of activity that responds to members' shared interests.The activities in question will differ in nature and in their significance both to the membership collectively and to individual members (Bennett, 2000).Furthermore, the scope, or range, of the programmes of activity of meta-organizations will vary.In this section, we examine how the make-up of the metaorganization's programme of activity influences the resource flows that can be expected.Berkowitz et al. (2022) discuss four types of meta-organizational activity, including (1) governance or co-management activities to facilitate interaction among members, (2) advocacy to ensure joint action, (3) boundary and category work to support the creation of identity and status, and (4) service production.These types of activity differ in terms of the requirement for involvement of member organizations in the activity and respective resource flows.

Nature of activity
Certain activities require participation by members.Members may have the necessary expertise, tools or technology to co-produce an activity, or they may only benefit if they participate.Whichever way it is considered, members' participation is necessary in these cases to make the activity meaningful, relevant or useful.Berkowitz, Bucheli and Dumez (2017) describe the activity of the Global Business Initiative, which includes regular workshops facilitating knowledge exchange among members.Although it may involve other resource flows, member resourcing, or indeed generated resourcing which compensates the member participation, will be essential.
Other forms of activity may not require member participation to make them meaningful, relevant or useful.The meta-organization may collect and analyse information about the environment, to produce, for example, industry sector reviews, government policy analysis and technology forecasts of mutual interest.These tasks, while they can be assigned to member representatives, are often assigned to meta-organization staff or to external experts.These activities may establish a flow of generated resources when sold and can be based on member resourcing, but more often depend on associational resourcing.
Finally, some activities become more meaningful, relevant or useful when members do not participate in the activity.For example, a meta-organization established as a regulatory or certifying body, or one with a significant monitoring function, must maintain a certain distance from the membership.Establishing a secretariat can safeguard trust in the fair dealing and messaging of the meta-organization in the eyes of both members and external parties (Berkowitz & Souchaud, 2019;Greenwood, 2002).These kinds of activity also often depend on associational resourcing, though when sold as a service may initiate a generated resource flow as well.
Meta-organizations may have to rely on resources in their environment to be effective enough to survive.Greenwood's (2002) research into EU associations highlights the importance of recognition as a legitimate stakeholder, access to policy processes, and an audience.A secretariat, acting as the guardian of the collective and long-term interests of members, can enhance the standing of the meta-organization and its members, and encourage an ongoing interchange between the metaorganization and external interests.In other activities, such as information production, the inclusion of recognized, independent sources of expertise and advice may be necessary for the activity to have traction.Those activities benefit from contributed resources.
Finally, when the meta-organization produces club goods or services of interest to parties beyond the membership, members may ask the meta-organization to generate resources, either by selling those goods or services or seeking donations or grants to enable development of such activities.The sale of broadcasting rights by FIFA is an example of a generated resource flow (Berkowitz et al., 2022).Actors in the wider environment might also benefit from a metaorganization undertaking certain activities (Ahrne & Brunsson, 2008;Provan, 1983).A good example is the benefit the European Commission gains from the existence of European Women's League, an organization fighting for equal recognition and opportunity (Karlberg, 2019), which it is therefore willing to fund.Such generated resourcing may supplement, or substitute for, associational resourcing.

Significance of activity
In addition to asking whose involvement is necessary for, or beneficial to, different types of meta-organizational activity, there is also the question of the significance of activities to members.There are two points to consider here: the significance to members collectively and to individual members.Provan (1983) argued that when the activities of a federation are important to its members, members will readily engage with, and may seek to influence and control that activity.Greenwood (2002) reports of one of the EU business associations recognized as a policy partner by the EU, that [its] 135 members and associate members. ..provide for a teeming structure of working parties and committees bursting with life.The expertise and authority of the senior industry figures involved in this work account for the high regard in which the organization is held.(p.14) Equally, Provan (1983) highlights how, when the issue is of limited importance, organizations still may seek management of their interdependence through federation affiliation, but they are unlikely to feel it necessary to become actively involved with one another for purposes of federation management.(p.82) In deciding whether proposals to initiate, continue or revive an activity should be supported, meta-organizations will consider whether the activity addresses significant shared interests and should therefore be included in the membership deal, or whether it is of limited or minority interest.In the latter case, it may be included in the programme of activity subject to extended resource engagement by those interested members -for example, in the form of fees for service, or additional member resourcing.Where activity is significant to interests outside the membership deal, it may lead to resource flows, whether contributed or generated.

Scope of activity
The range of activities undertaken by a meta-organization also influences resource flows.The broader the scope of its activity programme, the greater the chance that the meta-organization will mix activities of a different nature.All things being equal, the broader the scope, the less visible the contribution of resources (and difference in contribution) by member organizations.In contrast, when a meta-organization has a narrow scope of activity, we can expect it to be more important to members that fellow members contribute equally, or on a basis agreed to be fair.Greenwood (2002) observed such alternative settlements: Many large firms are happy to pay for a greater share of the costs involved in running the association in return for a greater degree of influence, and in many associations, there are unwritten, and sometimes written, rules in which votes and subscription fees are weighted.(p.15) When a meta-organization broadens the scope of its activity, there is more chance that members' interests will vary and that levels of commitment to any particular activity will differ.Member organizations can be expected to invest more in activities that are more significant to them, given a choice.The broader scope of activity also heightens the chance that some members will be unable or unwilling to allocate the resources needed to maintain the whole span of activity.That situation will likely provoke discussions about the future shape of the meta-organization and how a fair and equal division of costs might be guaranteed.Two types of response are possible.First, where activities allow an increase in generated resourcing, then activities that may otherwise be seen by members as unviable may be developed or sustained.Second, there may be changes to the structure of resourcing and the membership deal.Members might be asked to increase associational resourcing or to accept membership fees that differ according to the activities members wish to access.

Ideal Types of Meta-Organization and Meta-Organization Variation
The literature on meta-organization and closely associated concepts features three rather similar frameworks delineating contrasting ideal types.Ahrne and Brunsson (2008) highlight a sliding scale with networks at one pole moving through a weak meta-organization and strong meta-organization to individual-based organizations at the other pole.Provan (1983) distinguishes federations on a sliding scale, similarly starting with coalitions and participatory federations, through independent federation and mandated federation to owned systems.Provan and Kenis (2008) distinguish three types of goal-directed networks: participant-governed networks, lead organization-governed networks, and network administrative organizations.
All three frameworks start from whether the membership or a secretariat manages and realizes the meta-organization's activity.Furthermore, each framework identifies a distinct form in which a secretariat acquires substantial resources from sources outside the membership deal.Below, we review these ideal types using the resource-flow perspective.
The participatory federation (Provan, 1983), the participant-governed network (Provan & Kenis, 2008) and the weak meta-organization (Ahrne & Brunsson, 2008) share the principle that members are expected to play an active role in managing and producing the activities of the metaorganization, as well as in its governance.In some cases, the production of activity is constituted by and inseparable from participation -it is part of the member resourcing flow on which the metaorganization depends.When this extends to management of the activity, which is a function carried out for the members as a collective, questions arise about how to share that burden.As suggested earlier in the section on member resourcing, members have found different ways to divide the effort required for management: on a continuous basis, on a project-by-project basis, or by rotation.Members may delegate some decision rights to the management of the meta-organization, although only temporarily (Rothschild-Whitt, 1976).These participatory meta-organizations are weak because their high levels of dependence on members means they lack autonomy (Garaudel, 2020).This can potentially restrict their being recognized as a distinct organization (Bor, 2014), and their presence and status in their environment will depend on the meta-organization's members' willingness and ability to represent the collective interest (Greenwood, 2002).
In the independent federation (Provan, 1983) and the strong meta-organization (Ahrne & Brunsson, 2008), members agree to cede certain decision rights to a secretariat which Provan and Kenis (2008) refer to as network administrative organization.The members may, then, have little or no involvement in the management of the specified meta-organizational activities.Members may engage with activities and governance of the meta-organization, setting its goals and monitoring progress towards those goals.Each member will also assess implications for their own activities and the benefit they draw from membership.In these senses, the secretariat remains accountable to the members for its translation of collective decisions into activities that provide benefit.But it also becomes an entity entitled to act on behalf of the meta-organization, and this increases the meta-organization's strength relative to members.This type of meta-organization relies substantially on associational resourcing.Typically, membership fees provide the necessary resources for the meta-organization to hire its own staff, rent an office, and so on.Member resourcing is focused, then, more on members' engagement in governance and on participation in activities.
The meta-organization may access resources from non-members with interests in the metaorganization's activity.Indeed, Provan (1983) recognized the significance of generated resourcing flows for meta-organizational form, notably in specifying the mandated federation as a distinctive type.Ahrne and Brunsson (2008), too, considered strong encouragement by an external actor a reason to create a meta-organization, but did not develop an account of this type.Provan and Kenis (2008) discuss the lead organization-governed network whereby an external party or one of the members is the mandating party leading the meta-organization.In these ideal types, governance shifts from the members collectively to a third party (either an external party or one of the members going beyond the membership deal), which can influence, more or less strongly, the direction and types of activities of the collective (Müller-Seitz, 2012).Activities of the type are managed and delivered by a secretariat resourced by the third party.This ideal type no longer meets the essential meta-organizational characteristic of a voluntary association of equals and would therefore not qualify as a meta-organization.Less extreme forms of mandating, where governance remains in the hands of members, could qualify (Karlberg, 2019), though due to resource dependencies, the meta-organization would be open to influence by such a third party.The third party would also expect to receive specific benefits from the meta-organization which may put demands on its activities.For example, governments or business organizations sponsor meta-organizations as a convenient mechanism for consultation or coordination (Chaudhury et al., 2016;Karlberg, 2019;Valente & Oliver, 2018).
However, there are other options for third parties than resourcing a secretariat: resources could be channelled to members to compensate them for undertaking particular activities.Provan (1983) describes certain federations as allocating a pool of funding to members for defined services.Bor (2014) described European Commission (EC) funded R&D consortia, these illustrate how a metaorganization can secure a generated resource flow (e.g. an EC grant) which is allocated to members to compensate for the resources they invest in managing or delivering meta-organizational activity.A condition of the grant was that the meta-organization and its members were accountable to the EC, requiring members to report expenditure and progress towards set goals.The EC's governing role included the right to withdraw its resources or demand change if the meta-organization did not deliver the expected outcomes.
The above case confirms that rather than putting ideal types on a sliding scale, they are better seen as the cornerstones of a field or conceptual space that can accommodate and locate various forms of meta-organization (see Figure 1).The space is bounded by two axes: The vertical axis reflects where and on what terms the meta-organization acquires its resources, either within or from beyond the membership deal.This dimension is sensitive to the influence of third parties on the governance of the meta-organization and recognizes that the proportion and significance of the resources made available outside the membership deal can vary in meta-organizations.Second, the

Resourcing from within the membership deal "Member resourced, participatory meta-organization"
Participatory federation (Provan, 1983) Participant-governed (Provan & Kenis, 2008) Weak meta-organization (Ahrne &Brunsson, 2008) "Associational resourced, centralised meta-organization" Independent federation (Provan, 1983) Network administrative organization (Provan & Kenis, 2008) Strong meta-organization (Ahrne & Brunsson, 2008) "Externally resourced, centralised meta-organization" Milder version of Mandated federation (Provan, 1983) or Lead organization-governed (Provan & Kenis, 2008) "Externally resourced, participatory meta-organization" Identified in this paper, for example, described by Bor (2014) Variety of meta-organization horizontal axis is concerned with what we term structure of participation in the meta-organization: whether activity is managed and produced by member organizations, or by a secretariat.In metaorganizations, the extent of activity managed and delivered by member organizations relative to that managed and actualized by a secretariat will vary.Members may take full responsibility for the meta-organization's activity or there may be extensive delegation of responsibility and control to the secretariat.The space created by these two dimensions is depicted in Figure 1.Conceptualizing variation among meta-organizations in this way highlights the potential mix of resource flows and the different patterns of resource-flows that meta-organizations can achieve.In the bottom left-hand corner of Figure 1, members jointly manage and produce the activity of the meta-organization, utilizing the resource flow that we have called member resourcing.Blau and Scott (1963) saw that the members of mutual benefit associations were often content to allow an active few to undertake the organizing, but where meta-organizations establish a secretariat with its own staff and office, it is likely this will be relying on some associational resourcing.The more this secretariat takes responsibility for the activities of the meta-organization, the greater the commitment of associational resources required for allocation and use by the secretariat.As this pattern occurs, the more we move toward the bottom right-hand corner of Figure 1.The secretariat may serve members based on the membership deal alone, but both members and external parties may also see value in specific activities outside that membership deal.The more the meta-organization is focused on activities requiring resources from outside the membership deal, the more it moves upward in the space.In the top right-hand corner of Figure 1, the secretariat is no longer resourced by members through the membership deal.Rather, the meta-organization depends on resources generated or contributed by being able to provide benefits outside the membership deal.Instead of having a secretariat producing beneficial activity outside the membership deal, meta-organizations might also rely on the knowledge and capabilities of member organizations.The more member organizations participate in the production of goods or services using their own resources, and the more they accrue benefits from them, the more we move towards the top left-hand corner of Figure 1.At the extreme top left of the space, instead of a separate, dedicated staff, it is member organizations that govern, and that manage and produce activity.

Discussion and Conclusions
The concept of resource flow, and those of resource engagement and resource-flow profile provide a language through which to address the issue of resource acquisition by meta-organizations, this having received little prominence in meta-organization theory.In their initial account, Ahrne and Brunsson (2008) restricted consideration of resource acquisition to those provided by member organizations, suggesting that members would supply what was required.We have argued that this conception of resource acquisition is too limited, giving excessive emphasis to the closed, inwardfacing character of the meta-organization.Drawing on Pfeffer and Salancik's (1978) seminal analysis of resource acquisition and adapting it to the constitutional features of the meta-organization, we suggest four types of resource flow: member, associational, contributed and generated.This classification of resource flows uses what Floyd, Cornelissen, Wright and Delios (2011) term a bridging concept.The membership deal, usually a part of the meta-organization's constitution, distinguishes the resources that the meta-organization can expect to receive based on the (standing) agreement with its membership, from resources that the meta-organization may acquire beyond this membership deal.This concept reconciles Pfeffer and Salancik's (1978) focus on resource acquisition from external interests, potentially creating resource dependencies and vulnerability to external influence, and meta-organization theory's preoccupation with the membership of the meta-organization.The concept invites a different conceptualization of the meta-organizational boundary, which we discuss in more detail below.The membership deal is also likely to indicate the activities that the meta-organization can sustain based on the resources members make available.Three factors affect the development of resource flows, both within and beyond the deal: first, the nature of the activities; second, their significance in relation to members' and other actors' interests; and third, the scope of the meta-organization's overall programme of activity.In the discussion, below, we provide a summary of propositions in the form of a decision tree.We discuss the usefulness of this for practice, but also recognize the need for further research to strengthen our understanding of the factors affecting resource flows, and attention needed to explore how resource flows affect activity in return.Finally, the resource-flow perspective highlights the structure of participation of member organizations in the organization and management of the meta-organization's programme of activity as an important source of variation among meta-organizations.This insight, as we discuss below, should be helpful to future studies about agency and autonomy of the meta-organization.

Resource flow and meta-organizational boundary
Our focus on resource flows suggests that the meta-organization can be understood as having multiple boundaries.Following Santos and Eisenhardt (2005), this is to be expected.The boundary which the resource-flow perspective indicates is different to the membership boundary specified by Ahrne and Brunsson (2008) in two ways (see also Cropper & Bor, 2018).The resource boundary is narrower than the membership boundary: it highlights that member organizations are only partially committed to the meta-organization, as the resources they promise to make available to the meta-organization are constrained to the resource contributions collectively agreed.The need for additional resources can result in extending the membership deal such that all members are equally obligated to provide more resources.It can, however, also result in requests from certain members or external actors.In this case, the resource boundary is broader than the membership boundary, as resources not only derive from the membership deal but also from negotiations beyond that deal.These differences are visualized in Figure 2.
The resource boundary will be more challenging to identify, empirically, than the membership boundary, which can be identified using a list or register of members.Nevertheless, the resource boundary is salient to any nuanced analysis of interests in the meta-organization (Berkowitz et al., 2022;Wry et al., 2013).In reviewing the legacy and significance of RDT, Wry et al. (2013) commented on the fine analysis of interests and exchanges that would inform understanding of the tensions that resource acquisition might generate for an organization.For example, admitting resources from beyond the membership boundary may change the management focus of the metaorganization from an exclusive concern with the collective will of members to include the will of individual actors (see Karlberg, 2019;Karlberg & Jacobsson, 2015).Grandori (1997) comments on the significance of instituting a common responsibility among members to any third parties with which the collective engages in transactions.Indeed, she makes the distinction between associations and consortia in part on this basis.This common responsibility may function as a constraint on the activities and resource flows meta-organizations can sustain, as Ahrne and Brunsson (2008) suggest.The concept of multi-directional accountability (members to meta-organization, metaorganization to members, and meta-organization to others) developed by Bor (2014Bor ( , 2019) ) and discussed later by Berkowitz and Gadille (2022) could be a useful starting point for future research into the potential tensions and effects of resource flows.Other issues, such as reputational risk for member organizations resulting from resource flows, or the changes in the distribution of benefit among members and others when there are changes in activity and resource flow would also be interesting areas to explore.Pfeffer and Salancik (1978) would point to the processes by which dominant coalitions in organizations change activity to meet their interests more fully, and to changes in the dominant coalition itself, over time.In short, tracing the effects of either the use of a specific type of resource flow, or combination of resource flows, would be a fruitful avenue for future research.

Resource flow and activity
While previous research has highlighted that meta-organizations may differ in the type of membership composition (Bennett, 2000;Berkowitz et al., 2017;Saniossian, Lecocq, & Beaucourt, 2022) and may differ in what they do, their purpose and the activities they undertake or the role they play (see Berkowitz et al., 2022;Bor & O'Shea, 2022), the current research goes beyond this general distinction of type of member and type of activity or role.The focus in this research was on the difference made by resource flows.We highlighted that the preference of one resource flow over another is influenced by the nature of the activities, their significance for members' and other actors' interests, and the scope of the meta-organization's overall programme of activity as they affect the preference of one type of resources flow over another.The effects identified are summarized into the decision tree presented in Figure 3.
This decision tree can function as a helpful heuristic device for those working in meta-organizations.Those involved in meta-organizations can use it to determine the preferred type of resource flow considering the type of activity and the given circumstances in which such activity is being initiated.More generally, the tracking and recording of resource flows can be useful for those working in meta-organizations as it helps to clarify the different interests and potential sources of influence the organization is steered by.Similarly, the device can help meta-organizational research as it provides a quick overview of the different aspects affecting resource flows.The decision tree should be seen as providing an initial idea of factors.Future research could corroborate these factors as well as explore whether other factors also play a role in selecting the type of resource flow.
In focusing on the effect of activity on resource flow, this paper has not given attention to the question whether the availability of a particular resource flow may also affect the kinds of meta-organizational activities in return.König et al. (2012) suggest that meta-organizations are likely to be relatively stable, as meta-organizations tend to be inert, which would mean metaorganizations would have a set of activities that persist.Meta-organizations nevertheless change over time, not least in terms of their programme of activity, even if proposals for new activity are likely to be points of contention (Garaudel, 2020).What makes resource flows change, and what potential development paths meta-organizations may go through, is an interesting area for future research in which resource-flow perspective seems to offer insight.In addition, the role played by resource flows in the introduction of new activities, or in choices related to the continuation or discontinuation of activities of the meta-organization, deserves investigation.For example, future research could try to uncover whether or when resource offers beyond the membership deal result in the initiation of new activity and what happens to these activities when such resources are withdrawn.

Resource flow, member participation and meta-organization autonomy
There is a paradox in meta-organization theory.The theory holds that a meta-organization is constitutionally bound to its members and dependent on them for resources and, yet, must also have some autonomy and control over some resources to be recognized as an organization.As Garaudel (2020) argues, the relationship between the members and meta-organization may be more properly understood as one of mutual dependence, since the collective effort produces benefit for members that they would not otherwise be able to realize.The resource-flow perspective shows that it is possible to extend this argument a step further and thereby clarify where the impetus or force for associational activity comes from.There are two lines of argument that could be built.The first argues that the delegation of decision rights regarding the use and allocation of resources from members to the secretariat to ensure the management and the organization of activity, leads to the establishment of a unit that has or can develop a will of its own.This unit's interests may square with, but may potentially differ from, the immediate interests of the membership -for example, by emphasizing members' long-term interests (Greenwood, 2002).The second argues that the metaorganization's exchanges with organizations, outside the membership deal, has consequences for its ability to alter and extend members' articulation of their collective interests and thus engage agentively with organizations beyond the membership.Both lines of argument highlight the possibility of a greater sense of autonomy and agency for the meta-organization in the sense as suggested by Emirbayer and Mische (1998, p. 964): 'actors capable of formulating projects for the future and realizing them, even if only in small part, and with unforeseen consequences, in the present'.They suggest two mechanisms by which a meta-organization may develop into an organization with responsibilities and action impulses that are distinct from those of any of its members.Ahrne and Brunsson (2008, p. 132) observed that 'meta-organizations can best be understood as being in a transitional phase between a weak organization with strong members and a strong organization with weak members'.In addition to providing a means of exploring variations in metaorganization that are due to patterns of resource acquisition, the resource-flow perspective presented in this paper also provides some tools to help guide research in exploring why, when and how particular resource flows impact on this dynamic, and whether there are changes over time associated with the patterns of resource-flow we have discussed.

Final remark
The resource-flow perspective develops an account of meta-organizations that focuses on resources and activity.This may be particularly significant when the purposes of meta-organizations extend beyond that of identity and status work, emphasized by Ahrne and Brunsson (2008) and much of the subsequent inquiry.The resource-flow perspective also provides meta-organization researchers with additional conceptual tools and invites them to pose questions and develop research projects -theoretical and empirical -that extend and enrich our understanding of meta-organizations.

Figure 1 .
Figure 1.The ideal types and variations of the meta-organization.

Figure 2 .
Figure 2. The different boundaries in meta-organizations.

Table 1 .
The source and control over resources in meta-organizations.