Corps de l’article

A wide movement of inter-organizational restructurings has been going on for several years in the French non-profit field, and it has significantly increased lately (Tchernonog and Prouteau, 2019), resulting in higher concentration of the sector. The largest nonprofit organizations (NPOs) are becoming more powerful, while smaller ones are finding it increasingly difficult to access public funding (Eynaud and Tchernonog, 2019). To prevent this situation, NPOs are now trying to develop some restructuring strategies at the inter-organizational level.

In the literature, inter-organizational restructurings (IOR) refer to a series of more or less formalized collaborative practices between two or more NPOs, including resource pooling, partnership, mutualization, or mergers and acquisitions. Most of the empirical studies devoted to IOR among NPOs are located in the U.S. context, based on single or multiple case studies, and focused on mergers. They try to unveil the motivations that drive NPOs to restructure, to describe the decision and implementation processes, and to assess outputs of these changes, in quite a normative perspective leading to a series of operational advices. Notable exceptions must be mentioned, especially in the French context where several studies strive to exemplify the diversity of IOR practices and to highlight the existence of innovative IOR practices based on cooperation, which may constitute interesting alternative solutions to mergers (Fraisse et al., 2008; Richez-Battesti et al., 2017).

We draw both on the existing academic literature and on the general observation of IOR among NPOs in France, and we suggest to better depict the array of existing IOR practices, and further understand their motives and underlying logics. Recent analyses tend to give a rather negative view of those restructuring processes: they outline a weakening of the capacity for social innovation and for addressing new needs, in favour of a normalized response to specific needs that have been pre-identified by public authorities (Gardin et al., 2008). However, while groupings and mergers seem a priori far from the democratic and egalitarian functioning specific to the social and solidarity economy, some cases portray a specific attention to the question of cooperation, a prevalent concern in the field (Marival et al., 2015a). In this paper, we address two interrelated questions: first, what are the diverse forms of IOR that NPOs carry out, and what are the main drivers for their decisions? Second, in a context of increasingly strong constraining forces coming both from the market and from state, to what extent do those diverse forms of IOR mirror some specific features?

To answer this twofold question, we performed an online survey sent to a wide sample of French NPOs. First, we present a review of the literature devoted to nonprofit IORs, both in its empirical and theoretical dimensions. Second, we provide some insights on the empirical context of the French nonprofit field and on our research design and method. Our findings are elaborated in the third section, and discussed in the fourth and final section. Our research exemplifies a variety of types of restructurings between NPOs, along with a variety of motives for restructuring. We thus contribute to the nonprofit literature about IOR — stressing the specificities of IOR practices in the nonprofit, particularly in light of their relationships with public authorities, of the question of the hybridity they reflect and, more generally, of their cooperative dimension.

IOR Among NPOS: Literature Review, and Research Issue

Defining and Classifying IOR Forms

NPOs operate in tumultuous environments characterized by strongly constrained funding, increased competition, and greater demand for social services: in order to maintain their activity, NPOs undertake inter-organizational relationships, alliances, partnerships, restructurings, possibly including mergers. This phenomenon is linked to pressures from the environment and public authorities targeting budget cuts (Golensky and DeRuiter, 1999; Richez-Battesti and Malo, 2012), to a business-like tendency (Maier, et al., 2014), and to mimetic behaviors (DiMaggio, 1988). Mergers among nonprofit organizations (NPOs) have become more common in European countries (Abecassis et al., 2014; Grenier and Guitton-Philippe, 2011; Buckley et al., 2011; Tchernonog and Prouteau, 2019) as well as in North-America (Nonprofit Finance Fund, 2010; Blumberg, 2009).

The existing literature, however, remains unclear about what is meant by inter-organizational restructuring (IOR). Some alternative and close expressions coexist, such as “inter-organizational relationships” (that does not include mergers), or “inter-organizational collaborations” (that seems synonymous with IOR). Admittedly, even if mergers appear to be at the heart of US nonprofit scholars’ concerns, it does not preclude the existence of different types of IOR. Guo and Acar (2005) for instance deplore that the literature does not outline the specific characteristics of NPOs collaborations and hardly exemplifies the diversity of practices. They suggest a continuum from informal collaboration (information sharing, referral of clients, sharing of office spaces, and Management Service Organization) to formal collaboration (joint program, parent subsidiary, joint-venture, and merger). In the same vein, Proulx et al. (2014) identify eight models of collaboration between NPOs, also based on the degree of formalization and the degree of integration between the organizations involved. These two criteria, however, may be quite reductive of IOR practice, and “unlikely to capture enough of its multidimensional complexity” (Gazley, 2017, p. 1), especially because collaborative behaviours have to be analysed on multiple levels and at different points in time.

In our paper, we privileged the term “IOR” — for inter-organizational restructuring - considering that it is one of the most frequently used terms to refer to the wide spectrum of forms of collaboration in the NPO field and that, compared with “IO Collaboration”, it better highlights the increasingly constraining context where NPOs are currently operating.

Main Empirical Insights from the Literature on Nonprofit IOR: Motives and Effects

Most of the academic studies are performed in the U.S. context, based on empirical work, single or multiple case analysis, and consist in unveiling the major drivers or motives of IOR, in describing the decision and implementation processes, and in assessing their main effects, in quite a normative perspective.

Some scholars identified a set of motives that could explain IOR strategies in the non-profit field, referring to classical typologies developed to analyse profit-based organizations. In that vein, Trautwein’s typology (1990) is frequently mentioned in the literature: it distinguishes strategies resulting from managers’ rational choices, from a cognitive process of internal decisions, and from a reaction to an external event or change. This leads Pietroburgo and Wernet (2010) to identify two major perspectives to analyse NPOs mergers: the efficiency and rational choice perspective (when seeking synergies, strategic complementarities or a better strategic alignment) and the process perspective (when interpersonal dynamics, political logics, negotiation processes, and organizational alignment prevail). Benton and Austin (2010), for their part, suggest a somewhat different typology of motivations: improving the financial situation of at least one of the organizations; strengthening their visibility towards community and improving services. Harris et al. (2002) insist that it is rather a “web of complementary factors” that leads NPOs to engage in a merger, and that personal factors weigh just as much as more rational factors.

Finally, the effects of IOR processes appear to be contrasted. On the one hand, it may be “the death of an organization or the failure of a mission” (McCormick, 2001), even if several scholars point out the lack of research on cases of failed collaborative practices (Lundström, 2012). Correlatively, NPOs often encounter failures in their attempts to merge, due to a lack of anticipation in the decision-making process, or the difficulty to manage the integration phase and to predict long-term effects (Pietroburgo and Wernet, 2013). On the other hand, IOR strategies may be viewed as a positive alternative to dissolution, an opportunity for growth, a way to build organizational capacity and, even more, a way to improve the organization’s ability to serve its users and to fulfil its mission (Campbell, 2009; Benton and Austin, 2010; Norris-Tirrell, 2001; Yankey et al., 2001; Jones et al., 2017). For his part, Chen (2010) strives to integrate antecedents, processes and perceived outcomes of IORs in a single framework, and he shows that processes play as mediators between antecedents and outcomes, with a particularly significant role of resource sharing and trust building.

Alongside this rather “mainstream” literature, there is some research produced on the French context. In a comparative analysis of the French and Canadian contexts, Richez-Battesti and Malo (2012) confront different types of IOR with their underlying strategic motives. They identify three categories of strategies: growth strategies based on vertical integration and industry logic, autonomy maintenance strategies based on pooling or mutualisation, and resource mobilization strategies based on broadening internal or external stakeholders. In this last case, poles of territorial and economic cooperation (‘PTCE’) can open innovation perspectives by offering NPOs a regulation vector of the local economy that is distinct from administrative and competitive regulations (Fraisse, 2017b, 2017a; Fraisse and Gianfaldoni, 2017).

More generally, and despite its relative scarcity, research devoted to IORs between NPOs in the French context usefully highlights practices that are most specific to NPOs and most likely to generate social innovation, rather than focusing on the analysis of mergers from for-profit inspired perspectives. A number of reports have been produced on this point. To name a few, in her analysis of interorganizational cooperation in the cultural field, Deniau (2014) explicitly stresses the difference between mutualization — which is much about tools or methods, and cooperation — which refers to sharing a common project and acting or working together. He argues that the lexical diversity to speak of these interorganizational collaborations reflects the plurality of the phenomena and discourses, but also mirrors a high vitality of practices and of cooperative spirit in the nonprofit field, which can serve the general interest or even a willingness for social transformation, based on fundamental principles such as reciprocity, exchange, nonprofit solidarity and caring (Deniau, 2014). Marival et al. (2015a) also privilege the word “cooperation”, which they understand in a very broad sense, covering any form of groupings, partnership or pooling of resources, skills or activities for mutual benefit.

Underlying Theoretical Frameworks

The conceptual frameworks that dominate in the international literature about nonprofit IORs either refers to resource dependence theory, transaction cost theory, network theory or institutional theory, and are sometimes combined (Galaskiewicz and Bielefeld, 1998; Guo and Acar, 2005). Campbell (2009) first distinguishes ‘adaptation theories’ that refer to questions about the consistency between internal and external organizational dimensions: this includes resource dependence theory (IOR strategies aim to compensate the lack of available resources), political theory (organizations implement IOR strategies in order to get more power and/or to face a threat to autonomy), institutional theory (the main objective is to gain legitimacy towards stakeholders), and transaction cost theory (implementing an IOR is expected to decrease transaction cost and improve efficiency). Second, some theories are focused on the environmental dynamics and uncertainty, conceiving IOR as a way to better control external conditions that threaten performance or even survival, usually in a shared problem domain. Besides, in their analysis of 40 years of nonprofit collaboration research, Gazley and Guo (2015) and Gazley (2017) mention no less than three dozen of additional theories, most of them being oriented towards understanding human behaviors (cognitive and psychological frameworks).

All these theories are part of the classic “panoply” of frameworks for analyzing business strategies, without any particular attention to the possible specificities of nonprofits. Underneath the increasing IOR phenomenon, the question is about the choice of an economic model for NPOs, and most of the controversies are generally locked in a dual state-market approach that does not take into account NPOs’ specificities. Following Pietroburgo and Wernet (2010), scholars usually distinguish economic rational choice on the one hand, and political and personal factors, on the other hand, even if these two perspectives weight the same (Harris et al., 2002). Hence, the common opposition between mergers (close to rational choice) and cooperation (close to pooling of resources and political perspectives).

In the wide range of non-profit IOR forms, cooperation plays a crucial role. Interestingly, (Marival et al., 2015a) show the centrality of the social and political purposes of cooperation practices, as well as that of the associative project and of governance, which are frequently revisited before or during the IOR process, even in the most integrated forms such as mergers (Marival et al., 2015a). This refers to the willingness for social transformation and the sharing of a common project (Deniau, 2014). In our research, we will not use the term “cooperation” neither in its usual sense (inter-organizational relationships between independent organizations, hence excluding mergers), nor in its broader sense (referring to what we, following other scholars, designate here as ‘inter-organizational restructuring’ / IOR). We will rather privilege a conception of cooperation as a dimension or a feature that might more or less characterize any form of IOR, referring to NPOs’ capacity to maintain or develop a social and political stance linked to their core missions. As such, cooperation is usually associated with hybridization of resources, i.e. the ability to rally volunteer contributions beside public and market resources (Eme y Laville, 2006; Gardin, 2008).

On this basis, our general underlying argument is twofold: first, NPOs do carry out diverse forms of IOR, potentially driven by different motives; second, the cooperative dimension may be expressed to varying degrees in IOR, and may coexist with constraining forces coming both from the market (viability and survival constraints) or from the State (pressures by public authorities).

Method: An Exploratory Survey Among NPOs In France

To support our argument, we conducted a questionnaire survey within the entire French nonprofit sector. In the following section, we specify the main contemporary characteristics of NPOs in France, before explaining the methodology of our survey.

The French Nonprofit Field

The history of French NPOs is very long and deep-rooted, although the inclusion of NPOs in the French public sphere is quite a recent phenomenon (Archambault, 2001; Wagner, 2000)(Archambault, 2001; Wagner, 2000). Archambault (2001) explains that statism underpinned the central State’s “fight against local power in any form.” Historically, the State has either responded to its citizens’ needs or been expected to remunerate others delivering welfare services (including hospitals, asylums, and schools). Tchernonog and Prouteau (2019) estimate that France had 1.5 million operating NPOs in 2017, but that only 159,000 employ an estimated 1.568 million full-time equivalent staff (7.1% of the total French workforce), 56,8% of which work in the social and health-related sector, and are mostly funded by public resources. Staffed NPOs expend 88.2% of French NPOs’ combined annual budgets. Less than half (44%) of their funding originates from public resources and, more specifically, 20% from public subsidies (ibid.). Even if public subsidies are declining, they still appear as an important means for the French NPOs to get non-market resources. The volume of volunteer work in 2017 is 1.4 million jobs (full-time equivalent). 72.2% of this voluntary work is carried out in associations without salaried employees. Between 2011 and 2017, the number of volunteer participations increased by a little more than 30% (ibid.). The strength of the volunteering contribution can be conceived as a concrete expression of the concept of cooperative dimension.

More generally, the role of socio-political and institutional factors in the French nonprofit field is worth highlighting, even if it is part of a wider trend observed in many other countries. The question about IOR among NPOs is quite old because of a long history of relationships between NPOs and the French State. Two main factors have made the situation evolve: the political decentralization process gave more power to local authorities for funding and for assessing NPOs activities; the decreasing public grants for NPOs and the increasing public bids triggered a new competitive regulation that deeply transforms the interaction between NPOs themselves at the local level.

Questionnaire and Data Collection

The survey was designed after an exploratory phase made of twelve unstructured interviews with NPOs managers or chairmen that had implemented some form of IOR.. After identifying the most relevant issues emerging both form the literature and from the exploratory interviews, the questionnaire survey was structured around 6 main themes and a total of 52 questions. The questions are listed in table 1. A first draft of the questions was written by the authors and submitted twice to three professionals specialized in NPOs counselling and funding in order to improve their clarity for the intended audience.

Table 1

Structure of the questionnaire

Structure of the questionnaire

-> Voir la liste des tableaux

The survey was administered online and targeted French NPOs that had carried out or were in the process of carrying out a form of IOR. Since no specific database specifically linked to IOR could be acquired, the choice was made to submit the survey to a large sample of 20,000 NPOs representative of the French non-profit field in terms of activities. Sampling was based on data from Tchernonog’s study (2013). 770 answers were collected from NPOs among which 262 had conducted (63%) or were conducting (37%) an IOR operation and had coherently completed all the questions of our survey. The distribution of respondents by position within the organization is shown in Table 2. In most cases, they are executive managers (43,7%) or chairmen (25,4%), more rarely other board members (17,1%) or employees (13,1%).

Table 2

Function of individual respondents within the organization

Function of individual respondents within the organization

-> Voir la liste des tableaux

Data Analysis

Descriptive analyses were performed so as to characterize our sample of respondents and the diversity of IORs. Principal component analyses (PCA) were then carried out in order to group variables linked to the motivations and effects themes: PCA enabled us to reduce the complexity of the collected data into a few dimensions. In accordance with our research question and our objective of isolating the different logics at stake in these IOR operations, we also chose to conduct a hierarchical cluster analysis (HCA). More specifically, we relied on the Ward method, consisting in grouping cases into clusters such that the variance within a cluster is minimized. The HCA was performed using a set of variables linked to the different themes of the survey (motives, process, effects).

Our research method undoubtedly suffers some limitations. The survey leads us to collect declarative statements rather than facts. Besides, when it comes to IOR issues, it would have been useful to get answers from all the organizations involved in the IOR, but also from different stakeholders involved in the process (managers, chairmen, employees, volunteers, funders…). Besides, it is hard to assess the representativeness of our sample, since we do not have any detailed information about the whole population of non-profit IORs in France.

Results

In this section, we successively provide a general overview of the sample of respondents, their motivations for carrying out an IOR, the different forms of IOR identified, their effects, and their cooperative dimension. We then describe the typological analysis, along with the corresponding clusters, making it possible to identify different logics of IOR.

Descriptive Statistics: A General Description of the Observations

A First Overview of the Sample: Activities and Characteristics of the Partners

Our sample includes associations from all sectors of activity: health, social inclusion, culture, sport, education, etc. There is, however, a quite remarkable representation of organizations in health or social activities (see Table 3). While they represent 10% of NPOs in the general population (Tchernonog and Prouteau, 2019), they constitute 34,8% of our observations. Correlatively, associations of large size appear particularly represented among the samples (see Table 4). These two findings are related, as health and social sectors include organizations that are much larger than on average. On average, the share of public funding in the total budget of the organizations is 49.3% (details about the resources are provided in Figure 1). Finally, the respondent organizations are on average 40 years old (while the median age is 25).

Table 3

Activities of the respondent NPOs

Activities of the respondent NPOs

* Source: Tchernonog & Prouteau, 2019

-> Voir la liste des tableaux

Table 4

Budget of the respondent NPOs

Budget of the respondent NPOs

* Source: Tchernonog & Prouteau, 2019

-> Voir la liste des tableaux

Figure 1

Average distribution of the different types of resources in the total budget

Average distribution of the different types of resources in the total budget

-> Voir la liste des figures

Restructurings mostly involve partners with identical (42,4% of cases) or similar activities (52,3% of cases), in a horizontal logic (when activities are the same) or a vertical logic (when partners’ activities are complementary). On the contrary, IOR involving totally different activities remain very scarce (5,3% of respondents).

Four Main Forms of Interorganizational Restructurings

IOR may take a wide variety of legal forms: contract, management mandate, creation of a joint association (or integration within an existing group), creation of a common legal entity, partial transfer of assets, merger, etc. Those decisions, however, do not necessarily lead to legal formalization, and many cases involve informal cooperation and pooling (provision of premises, of computer equipment, of competences, etc.). These different forms may be distinguished according to the degree of legal formalism and integration of the structures involved, but also according to the objective of the operation. Hence, we retained four types of IOR (see Table 5):

  • Pooling resources: this may give rise to a convention or the creation of a group of employers, for example, but it often takes place without a legal framework (24,4% of cases).

  • Performing together a joint action, namely developing a project, organizing a festival, or setting up an advocacy campaign (18,3% of cases).

  • Creating a legal joint entity: when the partners develop an organization of which they are members. It could be a joint association, a cooperative, a federation, etc. (26,7% of cases).

  • Merging or transferring partial assets: when organizations boundaries are modified, in favor of an absorbing organization or a new entity that replaces previous partners (30,5% of cases).

Table 5

The four main forms of inter-organizational restructurings identified in the survey

The four main forms of inter-organizational restructurings identified in the survey

-> Voir la liste des tableaux

The first two categories are not based on a particular legal form, as they can be implemented with or without adopting a legal framework. A first difference between these different forms of IOR relies in the number of organizations involved, ranging from an average of 18.6 partners in a joint action, 9.7 in the creation of a common entity, 4.7 when pooling resources and 2.6 in the case of mergers. Regardless of the form chosen and the objective given to cooperation, our survey stresses the need to replace those decisions in a larger process. In 85% of the cases, indeed, the decision follows previous relationships between the partners (common action, co-members of a same federation, previous pooling of resources). Interorganizational restructuring is therefore a long-term process: the duration of the setting up of the restructuring is on average two years, ranging from 16 months in the case of simple pooling, to 29 months for mergers or when creating a joint entity (see Figure 2).

Figure 2

Duration of the process according to the form of IOR

Duration of the process according to the form of IOR

-> Voir la liste des figures

Four Categories of Motivations for IOR

Our data confirm the difficulties faced by NPOs, already highlighted by other studies (Tchernonog and Prouteau, 2019). Half of the respondents (47,4% of cases) mention that the decision to regroup follows a threat to the financial sustainability of at least one of the partners involved. However, a deeper analysis shows that IOR are part of multifactorial processes. From a dozen of variables assessing the motivations to initiate restructuring decisions based on Likert scales (see Table 1), we carried out principal component analyses (PCA). It made it possible to isolate a set of uncorrelated variables (principal components), revealing four main categories of motivations to restructure:

  • reaching higher operational efficiency, thanks to resource pooling;

  • developing new activities or geographically extending them, through strategic cooperation;

  • solving internal difficulties, especially for hiring employees, volunteers, or board members.

  • meeting direct or indirect requirements of public funders: the development of public procurements combined with decreasing public subsidies leads to new performance requirements for NPOs.

We observe large and significant differences in the importance of these motivations depending on the type of IOR (see Table 6). In line with the literature mentioned above, mergers are characterized by the weight of public funders demands as well as by the need to solve internal issues and dysfunctions. On the opposite side of the continuum, the implementation of joint actions aims primarily at developing activities. Resource pooling is rather expected to increase the operational efficiency of the organizations involved. And when the creation of a common legal entity is at stake, the objective is both to develop the field of activity and to meet public authorities’ demands.

Table 6

Forms of IOR and types of motivations (differences of means)

Forms of IOR and types of motivations (differences of means)

* differences of means (or proportions) statistically significant at the level of 0,001

** differences of means (or proportions) statistically significant at the level of 0,002

*** differences of means (or proportions) statistically significant at the level of 0,004

**** differences of means (or proportions) statistically significant at the level of 0,008

-> Voir la liste des tableaux

Human Resources and Organizational Dimensions of IOR Among NPOS

The survey questionnaire included several questions about the decision and implementation process, as well as the effects on individuals, on governance, and on organizational dimensions. We will more particularly focus here on the bivariate statistics aiming at testing a potential relationship between the 4 forms of IOR presented above those HR and organizational dimensions. Table 7 summarizes the main statistically significant means differences according to the type or form of IOR

Some of the results in Table 7 deserve being commented further. First, all the variables appear to take significantly different means between mergers and non-mergers sub-samples, indicating that the merger form has some specific characteristics and effects in terms of HR and organizational dimensions: even if quite low, the level of contest of the decision is higher, the degradation of social climate is higher, all the effects on governance are stronger (evolution or creation of a new board of directors, redefinition of the balance between executives and chairmen, improved effectiveness of governance bodies, better representation of all stakeholders), but management efficiency is more highly improved, the project of the organization is more frequently redefined and allows a better internal cohesion around this project. Besides, when IOR is a merger, the respondents more frequently identify some obstacles such as the fear of identity loss, the employees’ fear about their future and the risk of creating some conflicts when dispatching power within governance structures. Second, within the sub-sample of the three other forms of IOR (resource pooling, joint action, joint entity), we can also outline a few significant differences: when IOR is conceived as resource pooling, the effects are stronger in terms of social climate and work conditions degradation, but the process seems to lead to a higher efficiency of management mechanisms; the fear of identity lost is stronger when IOR takes the form of a joint entity. Nevertheless, HR and organizational dimensions and effects of IOR among the NPOs of our sample remain quite limited according to the respondents, and the overall assessment of the process appears quite positive.

Table 7

Forms of IOR and HR/organizational dimensions (analysis of variance)

Forms of IOR and HR/organizational dimensions (analysis of variance)

* differences of means (or proportions) statistically significant at the level of 0,050

** differences of means (or proportions) statistically significant at the level of 0,010

*** differences of means (or proportions) statistically significant at the level of 0,001

[…] In brackets in the “merger” column: degree of statistical significance of the differences of means between the mergers operations and every other restructurings globally considered

-> Voir la liste des tableaux

The Cooperative Dimension of IOR

The specificity of mergers is frequently linked to external pressures and a need to increase operational efficiency. Beyond these specificities, however, the cooperative dimension in IOR processes should be stressed. In order to measure the role of volunteers and its consequences in terms of IOR process and results, a new variable is created, based on the share of volunteer contributions in total salaried and volunteer resources in the organizations. This allows us to observe a negative correlation between the volunteering contribution and the degradation of the social climate, i.e. NPOs carrying out an IOR and relying on a higher proportion of volunteers have more favorable results in terms of social climate (see Table 8). A more significant volunteering contribution is also correlated to a lower level of contest in the course of the IOR process (see Table 9). Finally, whatever the form of IOR, the involvement of volunteers in the IOR initiative is significantly associated with a lower degree of contest and a smoother running of the operation (see Table 10). It could be argued that such results are actually due to differences in terms of size or types of activities of respondent organizations. Indeed mergers, which are linked to higher difficulties, are more frequent among bigger NPOs and among those involved in health and social activities; however, we found no significant relationship between the sectors or the size (based on the budget, the number of employees or the number of volunteers) on the one hand, and the existence of a contest or the degradation of the social climate on the other hand. Thus, it appears that one major criterion for successful and innovative IORs is the capacity to mobilize voluntary resources, regardless of the type of organization or IOR considered.

Table 8

Correlation between volunteering contribution and degradation of social climate

Correlation between volunteering contribution and degradation of social climate

*. The correlation is significant at the 0.05 level (2-tailed).

-> Voir la liste des tableaux

Table 9

Correlation between volunteering contribution and level of contest

Correlation between volunteering contribution and level of contest

*. The correlation is significant at the 0.05 level (2-tailed).

-> Voir la liste des tableaux

Table 10

Volunteer initiative and contest in the IOR process (Chi-square test)

Volunteer initiative and contest in the IOR process (Chi-square test)

The relation between the variables is statistically significant at the 0,03 level (2-sided).

-> Voir la liste des tableaux

Towards A Typology: Sector and Public Funding as Major Distinctive Dimensions

Method - The cluster analysis approach aims at bringing out some groups of individual observations (clusters) sufficiently homogeneous with regards to some key dimensions so that they can be considered as reflecting a common underlying logic. We performed a Hierarchical Ascendant Classification (HAC) procedure by using the Ward method. A critical methodological step consists in selecting the variables to be processed in the statistical procedure. 8 variables were chosen to run the HAC, based on our research questions and on conclusions drawn from the literature survey. 5 of them relate to the motivations for IOR. More specifically, we selected the 4 survey Likert-items in connexion with the strategic motivations (to what extent the IOR operation was driven by the objectives of a) Diversifying the activities or competences; b) Extending the geographical scope of action) and the public funding-related motivations (to what extent the IOR operation was driven by the objectives of a) Gaining leverage with public funders; b) Meeting to public funders’ demands). As for the operational efficiency dimension, we used the Principal Component Analysis (PCA) factor, which encompasses different items in relation with resource sharing and pooling: Sharing expenses or having access to a resource or job share; Pooling accounting, management, treasury resources; Pooling human resources (employees, volunteers); Pooling infrastructures or premises; Pooling purchases; Pooling IT; Pooling communication. Beyond the 5 motivation-related variables, 3 other variables were introduced for the HAC procedure: a sector variable in 10 categories; the type of restructuring (mergers vs. forms of IOR); and a dichotomous variable which distinguishes restructurings according to whether at least one of the NPOs involved was enduring a threat in terms of financial sustainability. Overcoming the threat to financial sustainability is often mentioned as a motivation for IORs (e.g., Benton et Austin, 2010; Tchernonog et Prouteau, 2019), but this threat can also be viewed as a contextual antecedent that reinforces other motivations pointed out in the literature (for instance, the need to gain access to more external resources is all the more critical when the survival of an organization is at stake). The presence of this contextual element also often reflects a more defensive approach to IORs.

Results - The HAC leads us to bring out four distinct clusters of relatively homogenous restructuring observations. These four clusters clearly distinguish one another along two broad dimensions. Firstly, they strongly differ in terms of sector representation. Most notably it appears that the health and social sectors are widely overrepresented in clusters 3 and 4, while there is no NPOs belonging to these sectors in clusters 1 and 2. Secondly, the clusters strongly contrast one another as regard to the overall weight of the public funding-related motivations for the restructuring. Indeed, this kind of motivation is strong in clusters 2 and 4 while it appears to have a lesser role in cluster 3 and even less so in cluster 1. The four clusters are schematically represented in the matrix below with regard to the sector and to the public funding-related motivations dimensions (see Table 11). Beyond these two discriminant dimensions, other specific features are also noted in the matrix. For example, restructurings in cluster 4 appear as significantly driven by strategic motivations, especially in a geographical extension perspective. Moreover, in cluster 4, the most integrative kinds of restructurings (joint legal entities, mergers) are overrepresented.

Table 11

A two-dimension matrix drawn from the cluster analysis

A two-dimension matrix drawn from the cluster analysis

-> Voir la liste des tableaux

Beyond the specific features of each cluster, several interesting patterns may be noted and some tentative conclusions can be drawn. First, the cluster analysis confirms the strong specificity of the health and social sector in the French NPOs landscape. Table 12 provides the main differences along this sectoral axis.

Table 12

Main differences along the sectoral axis of the matrix

Main differences along the sectoral axis of the matrix

-> Voir la liste des tableaux

Some significant differences can be observed among the restructurings in the sector in view of their underlying logic. Hence, in cluster 4, the restructurings seem to be much more motivated by strategic and public funding-related reasons, while in cluster 3, IOR seems mainly driven by operational efficiency motivations. As hinted by the previous remark, it should be noted that the different kinds of motivations are not independent ones: the operational efficiency motivations level is higher in the clusters where the public funding-related motivations are less significant. However, public funding-related motivations and strategic ones appear closely and positively linked. For example, when restructuring is driven by public funding requirements, our results suggest that one of the main purposes is to match the territorial scope of intervention of the NPOs in the field with the territorial level of the financing local public authority involved. This connexion between public funding-related motivations and strategic ones is less pronounced when one considers the two clusters with no NPOs from the health and social sectors (clusters 1 and 2).

When considering more specifically clusters 1 and 2 (where the health and social sectors are not represented), one can observe significant differences that mirror the differences between clusters 3 et 4. Thus, the notable weight of the public funding-related motivations in cluster 2 is similarly associated with more integrative forms of restructuring, lesser operational efficiency motivations and more reliance to public funding. In contrast, one can also note specific differences that are not or significantly less observable between clusters 3 and 4. The more important weight of public funding-related motivations in cluster 2 is associated with more intermediate territorial level (between national and local) restructurings and cases with threats over the financial sustainability of at least one NPO. Conversely, contrary to how cluster 4 differs from cluster 3, differences between clusters 1 and 2 are much less significant regarding the prevalence of strategic motivations. More specifically, the stronger weight of public funding-related motivations is much less associated with differences in terms of activities diversification and geographical extension motivations, compared with what is observable for the two clusters characterized by a higher representation of the health and social sector. Table 13 depicts in more details each of the four clusters.

Table 13

Detailed characterization of the 4 clusters

Detailed characterization of the 4 clusters

* differences of means (or proportions) statistically significant at the level of 0,050

** differences of means (or proportions) statistically significant at the level of 0,010

*** differences of means (or proportions) statistically significant at the level of 0,001

-> Voir la liste des tableaux

Discussion

We would like to sketch out three main points as a way to discuss our empirical results. The first one refers to the “big picture” that illuminates the diversity of IOR practices and forms, opening to a continuum between mergers and more innovative inter-organizational cooperation; the second point explores the role of public authorities in driving nonprofit IORs from a public-policy perspective with regard to the increasing prevalence of NPOs in the delivery of public services; the third one considers this role of public authorities from an organizational theory perspective with regard to the concepts of inter-sectoral organizational hybridity and hybrid public service organizations.

The “big picture” of nonprofit IORs landscape in France — Our study complements the recurrent surveys conducted in the nonprofit field and covering a wide range of variables (Tchernonog, 2013; Tchernonog and Prouteau, 2019), the more qualitative analyses (Marival et al., 2015a; Richez-Battesti et al., 2017), or research devoted to a particular sector (e.g. Deniau’s report on the cultural field, 2014). The questionnaire survey method has the merit of drawing the major trends in the IOR phenomenon in France, identifying some recurring characteristics and contingency factors, as well as grasping the respondents’ perceptions of the organizational effects. We identified a variety of forms of IOR along with different types of drivers that led organizations to group themselves into these forms, which echoes several studies that highlighted a continuum of IOR, mostly regarding the degree of integration of the structures involved and the degree of legal formalism (Guo and Acar, 2005; Proulx et al., 2014). Surprisingly enough, this continuum of forms (from joint actions to mergers) does not stand out as a significant explanatory variable of our final typology, except if we contrast mergers with other forms of grouping. Similarly, the motivations for IOR only appear as illustrative variables of the four clusters identified in this typology, except the public-funding related motivation. This contrasts with the typology suggested by (Marival et al., 2015a, 2015b), which shows two axes of differentiation in the 10 cases of collaboration between NPOs in the health and social sector: the degree of resources and governance integration and the purpose of the operation (economic, social or political). The authors emphasize, however, that the social purpose remains the prime driving force for all the forms of IOR, sometimes supplemented by a political aspect, and sometimes by an economic aspect. They also note, even in the economically-driven mergers, a central concern towards the respect for the fundamental values of the social and solidarity-based economy, the maintenance of a democratic governance and the renewal of the associative project while respecting the identities of each entity. Hence the potential dynamic of learning, creativity and innovation triggered by cooperation in those “legally hybrid forms”. (Marival, et al., 2015a; Richez-Battesti et al., 2017). Our findings do confirm this observation on a larger scale and on a multiple-sector sample: in highly constrained contexts, a strong pressure from public authorities and/or the existence of a threat to survival tend to favor most integrated forms of IOR such as mergers, but this is not incompatible with trying to redefine the project, with giving volunteers a central role and voice and, more generally, with pursuing a cooperative logics; in less constrained contexts, when NPOs spontaneously decide to implement IOR, they tend to favor less integrated forms in which cooperation is central.

The role of public authorities in driving nonprofit IORs — Our findings bring to light the role of public authorities as an important IOR driver in the French nonprofit field. This reflects the increasing propensity of welfare states to define themselves as purchasers and regulators of services provided by private and non-profit businesses (Evers, 2005). NPOs are now commonly seen as important providers of social services (Struyk, 2002; Evers, 2005; Bar-Nir and Gal, 2011; Guo and Zhang, 2013), public services (Guo, 2007) or government services (Burnley et al., 2005). The prevalence of NPOs in the delivery of social welfare (Dahlberg, 2006) and welfare services (Milbourne, 2009) is associated with marked processes of change in “welfare governance” (Bode, 2006; Bar-Nir and Gal, 2011) and shifts in the “welfare mix” (Evers, 1995; 2005; Bode, 2006; Seibel, 2015b) that have been observed in numerous countries and various institutional settings (Struyk, 2002; Burnley, Matthews y McKenzie, 2005; Dahlberg, 2006; Proulx, Bourque y Savard, 2007; Conaty, 2012; Guo y Zhang, 2013). Nonprofit service organizations (Hasenfeld and Gidron, 2005; Evers, 2005) and their role in public policy have been emphasized, suggesting that the increasing role of NPOs in the welfare economy (Bar-Nir y Gal, 2011) is closely connected with the new public management paradigm and a shift from government to governance, in the guise of public governance, network governance and policy networks approaches (Jessop, 1999; Bode, 2006; Brandsen and Karré, 2011). This shift is also associated with trends toward privatization, the devolution of responsibility for service delivery in many sectors from national to local governments, and the decentralization and reorientation of government functions ((Struyk, 2002; Evers, 2005; Conaty, 2012; Bar-Nir y Gal, 2011).

Empirical researches have explored many related issues, such as the types of public services that are most liable to be provided by NPOs (Dahlberg, 2006), the modalities of NPO-public authorities’ relationships (Struyk, 2002) and their implications in terms of organizational practices, accountability, quality of services or performance evaluation (Burnley et al., 2005; Milbourne, 2009; Conaty, 2012). These issues have been discussed in the specific French context, especially from a normative perspective. A radical overhaul of the relationships between public funders and NPOs has been advocated in French parliamentary reports (Bocquet, 2014; Blein, 2014; Morange, 2008; Warsmann, 2009; Langlais, 2008). Though many of their recommendations are yet to be implemented, these reports have signaled a culture change: encouraging public funders to take a more active approach and presaging new control and evaluation mechanisms. The main recommendations are twofold. First, requiring public funders to become effective partners to NPOs and to move away from a culture of “granting subsidies” to one based on “public procurement contracting” (Langlais, 2008, p. 38). Further, boosting coordination between the local and national political bodies and administrations is expected to increase productivity, reduce cross- and multiple-funding and competition between NPOs (Morange, 2008). Second, reports recommend retreating from classical ex-ante control approaches toward an evaluation focused on the content and outcomes of funded activities (Langlais, 2008; Morange, 2008). However, the lack of professional skills in NPOs and funders, an accounting-focused, narrow-minded view of control, have prevented from the implementation of a more partnership-based approach (Morange, 2008). These twofold recommendations have transformed the nonprofit field into a more competitive one and entailed a restructuring process among NPOs in order to survive by gaining a critical size. Besides, the last territorial reform in France, in reducing the number of administrative regions, further increases this need for restructuring. The impact of these territorial and political changes is also mirrored in our typology: the two clusters of IORs that are characterized by strong public funding-related motivations (clusters 2 and 4) are also those associated with more intermediate territorial level (between national and local) restructurings.

IORs and organizational hybridity — From a more theoretical perspective, the increasing role of NPOs as public service providers has been addressed in different ways: for example, public service providers NPOs typically illustrate the “contractual” (Coston, 1998) or “subcontracting” (Proulx et al., 2007) types of interface between the government and the third sector. However, a fruitful perspective is to be found in the concepts of hybridity and hybrid organizations, especially “inter-sectoral” (Evers, 2020) hybridity, i.e. hybrid forms mixing core structural elements from the three ideal-typical “domains of society” (Brandsen et al., 2005): the private/market, the public/state and the third/civil society sectors (Evers, 2005; Billis, 2010; Keast et al., 2006; Karré, 2012; Brandsen and Karré, 2011; Jäger and Schröer, 2014; Seibel, 2015b, 2015a). Some scholars have a third sector emphasis and propose related analytical concepts such as those of social service organizations (Evers, 2005), voluntary provider organizations (Bode, 2006), hybrid third sector organizations (Billis, 2010), NPO/public sector hybrid organisations (Conaty, 2012) or service-providing voluntary associations (Seibel, 2015a). It has been argued that hybridity and change are permanent features of organizations, and that the “third sector” is inherently characterized by fragmentation and fuzziness: (Brandsen et al., 2005) suggest that all third sector organizations are, in one way or another, caring organizations, providing services or goods with a “dual” public (collective) and private (individual) nature. Nuancing this view of hybridity as a permanent and inevitable characteristic of the third sector, Billis (2010) admits that hybrid organizations are ubiquitous, but he suggests distinguishing between “shallow” and “entrenched” hybridity in the third sector. The notion of shallow hybridity refers to the fact that some organizations have moved into hybridity in a rather gentle fashion, causing minor disturbances, and not calling into question their basic third-sector identity. In contrast, entrenched hybridity becomes a co-feature (Evers, 2020) of the organisation: in the third sector, it usually begins as a result of receiving private and public sector resources through grants, contracts and sales: at the governance level, the governing body may find itself under pressure to accept permanent government or private sector representatives in return for resources and influence; at the operational level, entrenchment arises when paid staff become dominant in the delivery of the operational work and when a management structure with several hierarchical levels is established. Significant resources have increasingly to be secured, through adaptation to political imperatives, and/or through the market principles of cost and price (ibid.).

In view of our results, NPOs in the health and social sectors present some distinctive attributes that allow to depict them as entrenched hybrid public service organizations: they are bigger in average, and the IORs in which they are involved are characterized by more integrative forms, much strongly driven by public-funding related motivations. Moreover, they are more associated with geographical extension motivations in relation with organizations operating at intermediate territorial levels rather than local or national levels, and with an improved management mechanisms’ efficiency, but also an increased fear of losing identity (See Table 11).

In more general terms, IORs involve important strategic decisions and, therefore, offer an interesting field of observations to investigate decision-making drivers in an NPO context. Our study shows that they are subject in different degrees to influential factors inherently connected to the public and the market sector. When IOR decisions originate primarily from external pressures grounded in public policy rationales (for instance, the need to match the territorial scope of intervention of NPOs with the territorial level of the financing local public authority involved), one may discern in it another significant indicator of entrenched hybridity in addition to the characteristic descriptors specified by (Billis, 2010).

The inter-sectoral organizational hybridity perspective has undoubtedly some connection with alternative conceptualisations of hybridity, for example in terms of institutional logics (Battilana et Dorado, 2010; Pache et Santos, 2013; Battilana and Lee, 2014; Skelcher and Smith, 2015), but also with the “Polanyian hybridity” approach (Roy y Grant, 2020): the analytical focus is on the “triadic” structure of society (Roy & Grant, ibid.), with the principles of reciprocity (commonly via household/community/civil society), redistribution (most usually via the state), or exchange (via the market). The two perspectives are rather complementary, and the Polanyian approach rooted in the political and sociological (critical) tradition, whereas the inter-sectoral hybridity perspective is anchored in organization theory, including works with a central focus on hybrid NPOs (Evers, 1995, 2005, 2020; Brandsen et al., 2005; Billis, 2010).

Conclusion

Our main objective was to better understand IOR practices in the French third sector by grasping the plurality of their forms and motives, with a focus on the cooperative dimension. Our findings show that the market orientation cannot fully explain the diversity of IOR practices and corroborate the centrality of the social and political perspectives in IOR. The cooperative dimension is expressed to varying degrees in IOR, and coexists with different forces coming from the market or from state. Admittedly, IOR potential benefits that may be obtained from a bigger size appear as a frequent motivation for adopting an integrative approach. Restructurings in the nonprofit field, however, portray some distinctive features. One crucial point, more specifically related to mergers, is the role of market mechanisms in the process of bringing together different entities that were previously independent. In fact, two other drivers may be identified when it comes to understand why some independent entities decide to voluntarily give up their autonomous status. The first one is the institutional pressure by public funders, which is all the more likely to be influential when NPOs are predominantly financed through public funding. The second driver is related to the threats in terms of financial sustainability that NPOs may be facing. In some cases, mergers appear as the only way to ensure NPOs’ sustainability. This does not mean, however, that NPOs are not likely to cooperate in the absence of public funding constraints or threats in terms of financial sustainability. But NPOs may spontaneously prefer other form of restructurings consistent with the conservation of their autonomy and their identity. From this point of view, the large variety of legal structures of French NPOs can be understood as a specific response to the need for NPOs to act collectively, while desiring to maintain their independence and preserve the will of their volunteers. The logic of cooperation is supported through the action and the commitment of volunteers. By participating to the NPO’s governance, volunteers have a crucial influence on IOR: when volunteers, board chair or board members are the initiators of IOR, the level of contest is lower, and we can assume that, in this context, NPOs keep in shallow hybridity. This highlights the importance of non-monetary resources and of volunteers’ work in NPOs. NPOs should be reflexive about their own history before engaging in a functional restructuring process, and take into account their community-based orientation. By doing so, it is possible to have a long-term view on NPOs and to develop a more comprehensive approach. Because each NPO’s mission engages a community, the restructuring decision can be part of critical debates with all stakeholders, and NPOs can claim their difference and gain legitimacy.

One limitation of our research lies in our deliberately specific understanding of the concept of “cooperation” in the nonprofit field: the capacity to maintain or develop a social and political stance linked to the core mission, and the hybridization of resources (Eme y Laville, 2006; Gardin, 2008). In contrast, strategic management scholars more broadly refer to the idea of collaborating towards a common goal (e.g. Child et Faulkner, 1998). We strived to avoid any conceptual ambiguity in stressing our definition of cooperation. On the empirical side, our exploratory quantitative survey calls for getting qualitative insights in order to further explore the decision and implementation processes in different IOR contexts. Another line of research would be to set up a comparative study between several countries, including countries with different modes of public regulation: we could draw on the classification suggested by Archambault (2017) who identifies five clusters within the European model of government-nonprofit relationships or on Dor’s typology of socio-economic models of the third sector in Europe (Dor, 2020). We could also propose to direct the study towards a North/South comparison: most of IOR research in the nonprofit field gives little to no room to issues such as domination, class, solidarity economy, social movements, or popular economy (Eynaud et al., 2019). We claim the need to open new avenues by broadening IOR scope to the global south and by providing a better understanding of international pluralism.