Researching the political economy of adult learning systems. Theoretical amendments and empirical findings

The article relates to different approaches which emphasize a multi-level perspective as well as the institutional conditions that shape adult education. With reference to both the political economy approach and the governance perspective, it is outlined how these more general approaches can be theoretically amended by neo-institutional ideas especially when considering the institutional conditions that shape adult education at an organizational level. Based on a re-analysis of interview data, the findings provide first empirical indications of how institutional conditions of adult education organizations (e.g. the actors involved and their institutionalized relationships of interdependency) can be further specified when neo-institutional ideas are taken into account.


Introduction
The year 2016 marked the 200th anniversary of the beginning of comparative education: it was in 1816 and 1817 that Jullien de Paris published a questionnaire which is considered as the foundation of comparative education as a discipline. Field, Künzel and Schemmann took the anniversary as an occasion to reflect on the development of international comparative adult education (Field et al. 2016; see also Field et al. 2019). Taking account of globalization tendencies and the erosion of the nation state they asked the provocative question: "Is the chapter of comparative adult education research now closing?" (Field et al. 2016, p. 128). This question and the ensuing reflection generated a vivid debate (Schemmann 2017;Fejes and Nylander 2019), which featured a variety of aspects and points of view. One answer, even though not directly in response to the original question, was given by the publication of the book "Political Economy of Adult Learning Systems" (Desjardins 2017). By developing a distinct theoretical approach and applying it to the analysis of adult learning systems in eight countries, Desjardins answered the question with a clear "no". His study of the political economy of adult learning systems made an important contribution to the debate by highlighting to what degree adult education is embedded in contexts of economic and social institutions.
However, when analyzing the theoretical approach of political economy, it becomes obvious that the organizational level is hardly taken into account. This then is the starting point of the present paper, which focusses on two aspects: On the one hand, it intends to make a contribution to the theoretical notion of political economy of adult education systems by analyzing the institutional environment of organizations working in adult education. Hence, we introduce neo-institutionalism as a complementary theoretical approach that serves to open up perspectives for empirical analysis and operationalization. Furthermore, the paper applies this newly developed approach and by way of example, carries out an empirical analysis of public adult education organizations in order to prove the added value of this theoretical amendment.
The paper starts off with a presentation of the theoretical framework, i.e. the notion of political economy of adult learning systems. When focusing on the organizational level, we will discuss the theoretical perspectives of educational governance and neo-institutionalism. The third section will outline the methodical design of the study from which the data is taken and re-analyzed. The research question for the re-analysis is also developed here. The research findings will be presented in Sect. 4. The paper will conclude by discussing the contribution these findings bring to the debate on political economy of adult education systems and by highlighting further research perspectives.

Exploring and extending the theoretical framework 2.1 Approach of political economy of adult learning systems
A political economy of adult learning systems is rooted in the idea that adult learning systems are embedded in the characteristic context of economic and social institutions. This regime of social and economic institutions is not to be understood in a static way: "The level, distribution and diversity of adult learning in a given country is the product of ongoing sociopolitical negotiations in interaction with the prevailing structural and public policy frameworks that surround adult learning" (Desjardins 2017, p. 21). The nation-specific institutions have a profound impact on existing adult learning systems. Whereas Harney's system theory based approach understands adult education as a blank template that acquires both distinct structure and function when connecting to systems like the employment system, the vocational training system or the general and liberal education system (Harney 1997), the political economy approach sees interactions, or "institutional complementarities", between institutions in the sense that "the presence (or efficiency) of one institution increases the returns from the other" (Hall and Sokijize 2001). Desjardins highlights (1.) the education system, (2.) the world of work and (3.) the broader economic, social and cultural system as relevant systems.
The education system is considered core since it is closely linked to the adult learning system: types of formal adult education, second-chance education, and vocational training programs are directly connected to the initial education system (Desjardins 2017, p. 22). What is more, the education system also has a direct impact on adult learning since " [...] adults in better jobs are more likely to attract further investment from their employer in their opportunity to learn over their lifespan" (Desjardins 2017, p. 22).
Adult learning is also perceived as closely linked to the world of work, with the labor market and the employment system as important points of reference. Desjardins refers to the involvement of the private sector as both provider and consumer of adult learning as well as the state with active labor market strategies and labor market programs (Desjardins 2017, pp. 22-23).
Finally, the adult learning system is also seen to be embedded in the broader economic, social and cultural system. "The wider structural relations in society between state-market, state-civil society and market-civil society and their underlying institutions shape the prevailing structural conditions of authority and power, choice and opportunity as well as the distribution of resources, and in turn interact with, and impact, the nature and functioning of ALS [Adult Learning Systems]" (Desjardins 2017, p. 23). Thus, adult learning systems relate to social policies but also to broader state measures of social welfare, benefits, taxation, and redistribution.
Applying this approach of political economy of adult learning systems, Desjardins developed a framework for analysis and subsequently analyzed key differences and similarities between adult learning systems of eight countries. The analytical framework included aspects relating to governance, financing and service structures. Additionally, the study took participation rates based on PIAAC data into account (Desjardins 2017). This refers to a smaller strand of research which analyzes the impact of institutional settings on adult education systems in order to explain differences in participation. The basis for this research was provided in the "bounded agency model" developed by Rubenson and Desjardins (2009). That model proceeds from the assumption that welfare state regimes affect the individual's chance to take part in adult education: it is claimed that by fostering structural conditions that are relevant to participation, and by designing targeted policy measures, both structurally and individually based barriers can be overcome.
All in all, Desjardins' study provides a profound, internationally comparative analysis of policies and structures that foster adult learning, based on a differentiated database. At the same time, at least two aspects are worth highlighting when taking a critical look at Desjardins' study and his approach of political economy of adult learning systems. First, this approach mainly focuses on structural elements and aspects which are external to the organizations, and thus neglects to account for the contribution of adult education professionals in providing adult learning and shaping the adult learning systems. Secondly, the approach hardly takes the level of the organization itself into account. We will below work particularly on this second shortcoming and discuss multi-level perspectives on adult education that include the organizational level. What is more, we refer to neo-institutional ideas as a promising and complementary theory to introduce to the political economy approach.

Organizational level as an additional theoretical element of political economy of adult learning systems
There is a general consensus in debates of adult education that an exploration of adult education cannot be reduced to one level or dimension but needs to be viewed from a multi-level perspective in order to grasp its complexity (e.g. Schemmann 2015). We will subsequently refer to multi-level approaches from both a didactical K and a governance perspective before exploring neo-institutionalism as an approach from organizational theory. The multi-level perspective was introduced from a didactical point of view by Siebert (2000). Following on from the work of Flechsig and Haller (1975), Siebert developed a concept of didactical action which differentiates five levels. At the first level, he highlighted educational policy as having didactical implications for adult education. As examples, he referred to dedicated financial support for certain adult education programs, or regulations for the certification of adult education organizations. Siebert identified the organizational level as the next one, acknowledging didactical implications arising from cooperation or competition between organizations, as well as the structure and the overall goal of the organization (Siebert 2000). Next, departmental planning of the program, e.g. as regards consecutive or sequenced courses, has an impact on the didactical dimension. Last but one, the level of seminar planning has didactical implications since it makes a difference if initial tests or final tests are introduced or what kind of learning material or media are chosen. Finally, and most obviously, the micro-didactical planning of the teaching situation has implications since the social situation, the teaching methods used or the opening situation need to be considered (Siebert 2000).
Schrader, in turn, developed a multi-level model of adult education focusing on governance rather than didactical aspects (Schrader 2010). Schrader noted that there is little discussion about governance of adult education in Germany, most notably an absence of models on governance which take account of the diversity of actors, structures and processes involved. Consequently, he developed the "Framework Model of Multilevel Governance in Adult and Further Education" (Schrader 2010, p. 46).
As can be seen in Fig. 1, Schrader differentiates five levels, placing the level of teaching and learning processes at the center of the model. This level is "[...] understood here as the level of provision, utilization, and effect of learning opportunities" (Schrader 2010, p. 48). At a second level, Schrader refers to organizations as collective actors. Their function is to guarantee time-related, factual and social conditions  (Schrader 2010, p. 45) for teaching and learning processes, delivered by individual employees engaged in management, planning and administration (Schrader 2010, p. 50). A third level is defined as the immediate environment of adult and further education organizations, which includes various actors such as chambers, associations of providers, accreditation agencies or quality management agencies as well as statistical offices and scientific institutes. The fourth level is dedicated to state actors and state action. In the case of Germany, this involves the central state, the federal states (Länder) as well as the municipalities. These actors have the authority to both regulate adult education and to allocate financial resources. Finally, Schrader also takes international, supra-and transnational actors into account. Thus, the fifth level comprises the EU, OECD, UNESCO or the World Bank, who provide policies, e.g. on lifelong learning, or statistical monitoring reports and, as in the case of the EU, have enormous financial resources to fund programs and projects.
Another approach specifically focusing on governance, known in the German discourse as "Educational Governance" (Altrichter et al. 2007), was developed in school education and higher education research first before being applied to adult education as well (Schemmann 2015). Since then, a considerable body of work has been produced on the subject (see most recently Langer and Brüsemeister 2019). The aim is to understand "[...] how regulation and performance of school systems is achieved, sustained and transformed under the perspective of coordination of action between various social actors in complex multilevel systems" (Altrichter 2010, p. 148).
The quote above emphasizes that coordination of action is one important aspect affecting interdependencies of both collective and individual actors. What is more, this perspective is also interested in mixed or hybrid types of coordination, which suggests that different forms of coordination of action do not only co-exist but also interact and influence each other. From this perspective, there is not one single actor of relevance but a multitude of actors, which therefore makes the constellation of actors the focus of analysis. Furthermore, the educational governance perspective also assumes that stakeholders' actions are based on institutions or an institutional system of rules. The function of this institutional basis is certainty in decisions (Altrichter 2015). Finally, the educational governance perspective considers complex social systems like the school or the adult education system also as multilevel phenomena (Altrichter 2010, p. 150).
When relating the educational governance perspective and the political economy approach, it becomes obvious that both perspectives make a reference to the environment of the adult learning systems and the involved organizations. However, within the political economy approach the reference remains rather general as regards the differentiation of the environment in distinct system-realms. Within the educational governance approach, it remains general regarding the interdependency of various actors of the environment. Thus, we introduce the neo-institutional concept of environment for further specification of the term.
Neo-institutionalism offers a way of specifying the institutional foundations of the environment which is of relevance for the adult education organizations. Focusing on the rule-related, binding and regular elements of the environment, the approach thereby specifies the spheres of environment for adult learning systems identified K by Desjardins and the interdependencies between actors from the educational governance perspective.
From the neo-institutional point of view, organizations mirror their environment (Meyer and Rowan 1977). To ensure their sustainability in general, in terms of their internal structuring (e.g. Herbrechter 2018) as well as their external representation in particular (e.g. Hartz 2011), adult education organizations orientate towards institutional terms of reference and institutional expectations of their environment (for the educational science perspective on neo-institutional ideas, see also Herbrechter and Schrader 2018;Klingovsky 2016;Koch and Schemmann 2009;Kuper 2001;Kuper and Tiehl 2018;Schaefers 2002Schaefers , 2009Schemmann 2016;Tippelt and Lindemann 2018).
The concept of environment within neo-institutionalism was further elaborated by DiMaggio und Powell as "organizational field": "By organizational field, we mean those organizations that, in the aggregate, constitute a recognized area of institutional life: key suppliers, resource and product consumers, regulatory agencies, and other organizations that produce similar services or products" (DiMaggio and Powell 1983, p. 148). Thus, an organizational field is made up of by a group of organizations that are closely interdependent. These interdependent relations can take various forms, e.g. direct or indirect, hierarchical or non-hierarchical, and can exist between functionally different or similar organizations; they also convey expectations of appropriate and rational structuring of the organization (DiMaggio and Powell 1983). Even if they are part of the same overall social context, organizational fields can differ from one another in terms of how to design and structure modern organizations appropriately. In Scott's understanding (1994), organizational fields primarily vary according to: (1) the jurisdiction of belief systems (i.e. what is mutually recognized as the guiding basis of action), (2) the nature of governance systems (i.e. which forms of governance are considered to be effective), and (3) the degree of structuring within the field (i.e. density of interaction as well as awareness of the existence of other organizations) (see also Walgenbach 2014, p. 301;Greenwood et al. 2013, p. 3). Furthermore, DiMaggio and Powell make clear that the structure of an organizational field has to be defined on an empirical basis (DiMaggio and Powell 1983, p. 148).
All in all, neo-institutionalism complements the political economy approach by firstly emphasizing the organizational level and secondly, by specifying the various spheres of environment surrounding adult learning systems. The concept of organizational field as well as definition criteria help to empirically analyze the environment of the organizations as parts of the adult learning systems and offers an option for operationalization of research. What is more, the approach allows for an empirical analysis of which actors in what specific environment are of importance to which organization.
Below, we intend to apply this perspective as an example by re-analyzing data collected in the context of a study on the changes in governance of public adult education organizations. In addition, we further develop the research question of the study entitled "Governance structures in public adult education organizations (GLOW)", for which the data was gathered originally, in order to allow for contextualization.

Empirical design
The paper draws on the empirical data of a study which was carried out with support of the German Research Association (SCHE 585/2-1; DO 746/3-1). Its opening premise was the fact that several authors referred to a new governance regime within publicly funded adult education, taking account of changing actors as well as new instruments of governance and new levels that are of importance (Schrader 2008). However, a look at the current state-of-the-art research reveals predominantly descriptive compilations of the empirical changes-on both system and organizational levels. Empirical analyses are rare, except an analysis of the effects of quality management (Hartz 2011), an analysis of organizational development in view of extended network structures (Feld 2011), or an analysis of program planning against the background of new economic needs within an organization (Dollhausen 2008;Nuissl and Dollhausen 2011).
There are no studies to date which take a detailed look at this new governance regime. Even though there are empirical indications of changes, there is very little analysis around specific characteristics, relevance and effects of new governance mechanisms on the organizational level. What is more, models to explain the governance of adult education organizations are missing altogether.
The GLOW study picks up on this gap and focuses on the following research questions: How can the new governance regime of public adult education organizations be characterized? Which forms of coordination are of relevance? What are the consequences for decision-making and the legitimacy of decisions made by program planners in the organization? What are the consequences for the output of the organization?
As indicated in Fig. 2, GLOW aims to use case studies to answer the questions above and to explore how a new governance regime can be conceptualized. In terms of theory, the study leans on the perspective of educational governance (Kussau and Brüsemeister 2007;Altrichter 2015) combined with neo-institutional elements and reference to the theory of conventions (Boltanski and Thévenot 2007). K In order to identify cases for the study, a two-step-cluster analysis was carried out using the data of the VHS-Statistik (Statistics on Adult Education Centers) from 2005 and 2015 (Engels 2018). All in all, six clusters were identified out of which ten cases were drawn. For the organizations involved, three to seven episodic interviews were carried out with heads of the organizations and program planners. Episodic interviews are characterized by both open impulses for narration as well as semi-standardized questions (Flick 2019). Overall, 41 interviews were collected. The interviews were coded by two independent coders (Cohens Kappa κ = 0.68) and analyzed by method of content analysis (Mayring 2015).
As we have argued before, introducing neo-institutionalism into the theoretical approach of political economy of adult learning systems helps identify the actors within the systems (or the actors of organizational fields within systems) and the constellation of actors. In order to apply this theoretical perspective empirically, we re-analyzed the GLOW data for the purpose of this paper, focusing our analytical attention towards actors in the environment of publicly funded adult education organizations. For the re-analysis of the collected interview data we focused especially on the following research question: Which actors do leaders and program planners of public adult education organizations relate to as important in their environment, and how do they describe the interdependency relationships with these actors? Following the method of qualitative content analysis as defined by Mayring (2015), we analyze interviews with leaders and program planners as the key persons for the analysis of relevant actor constellations (Kieser and Walgenbach 2010, p. 167). Within their everyday working routines, they engage in frequent social networks exchanges and interact regularly with different environmental actors who are of relevance for the organization.

Findings
We will subsequently report the findings resulting from the re-analysis of data collected in the context of a study on the changes in governance of public adult education organizations. As mentioned above, our intention is to identify relevant actors in the environment of public adult education organizations and to analyze their relationship with these actors. For the report of our findings, we assigned the actors identified to the three relevant systems outlined by Desjardins, i.e. the broader economic, social and cultural system, the world of work, and the educational system. As regards organizations that are similar or alike, we will only point out this category when we find organizations in the environment that are indeed alike. As the sample of the GLOW study consists of publicly funded organizations, we assume that powerful actors of the broader public social system (e.g. political actors) are of special relevance. From a theoretical point of view, we expect, in terms of results, that all outlined forms of interdependence relationships (see Sect. 2.2) can be empirically found apart from direct or indirect hierarchical relationships between functionally similar adult education organizations (since they are theoretically possible but not empirically probable).
As regards the systems involved, we have to point out that most actors identified as important for the public adult education organizations could actually be assigned to the broader economic, social and cultural system. State actors like the municipal councils, mayors as well as the district government were seen as extremely important. The relationship towards them can be described as hierarchical and direct. A crucial point of reference for shaping the relationship with these public actors is the law.
And EVERYTHING we do is ALWAYS-and it must not be in a different way, is always defined by law, which is our basis. (Interview Neustadt, Mr. Kronberg) 1 [...] this morning I had a Jour fixe with our Lord Mayor and we discussed all relevant issues, especially budget questions, financing but also important topics for our program. (Interview Feldberg, Mr. Müller) The second quotation also shows the topics which are under the influence of the municipalities. In addition to budget issues, questions about permanent and nonpermanent staff are mentioned in other interviews as well as topics for the program.
At the federal level the Federal Office for Migration and Refugees (Bundesamt für Migration und Flüchtlinge, BAMF) is pointed out as an important actor. This certainly reflects the development of migration and numbers of refugees in Germany over the last few years and the fact that public adult education organizations contribute quite a bit to efforts of integration. This relationship can also be characterized as hierarchical and direct.
If the BAMF says we will only have permanently employed teaching staff in the future, then we will employ permanent staff, no matter what my treasurer here in the city says. Or we won't offer integration courses any more. (Interview Arlingen, Mr. Marques) The European Union has to be mentioned in this context as well. Both the European Social Fund as well as Erasmus+ are mentioned by interviewees. The relationship can be characterized as hierarchical and direct. In general, these actors are important for the financial sustainability of the organizations concerned: The third-party-funding sector is by far the strongest in our organization. [...] It means we have one third subsidies from the state, one third project funds and one third program funding contributions from participants. But the program contributions, I do not know how much it really is now, very marginal though. [...] And that is why the project funds, which is my area actually, is so important. (Interview Eigelsheim, Ms. Gimsen) Other important actors are professional associations such as the Federal Association of Public Adult Education Organizations (Deutscher Volkshochschulverband, DVV) or the State Associations of Public Adult Education Organizations (Landesverbände der Volkshochschulen, LVV). They often provide teaching material and take the lead in the representation of interests. The relationship can be described as non-hierarchical and direct: Well, and in the broadest sense the state associations have expectations, and the federal association is totally abstract to me, I do not feel obliged to them. The state association in Munich is relevant for us. (Interview Arlingen, Mr.

Marques)
The federal association has comprehensive curricula and also guidance materials for trainers as well as guidance for the processes, definitions of content, of what works and of what does not work. (Interview Feldberg, Ms. Gramberg) Working groups or local study groups also feature in this context. Finally, quality management organizations are mentioned as well since these organizations have a great impact on the internal structures of the adult education organizations. The relationship can be characterized as hierarchical and direct.
It does not come as a surprise that actors of the "world of work" are not mentioned prominently by the interviewees. Public adult education organizations are mainly committed to programs of general and liberal adult education. However, some organizations also offer programs in vocational adult education and training. Thus, the employment office is mentioned as one central actor in this regard and the relationship is described as hierarchical and direct. What is more, actors like chambers of commerce feature in the interviews. Here the relationship is characterized as direct if they are involved in projects, or indirect if they are involved in a campaign by the lord mayor's office, for example. In both cases the relationship is non-hierarchical.
Finally, actors from the education system are referred to as well, schools, however, only very rarely. If at all, it is vocational colleges that the adult education organizations have cooperated with. As such the relationship is non-hierarchical and direct. The interviewees referred to other adult education organizations more often, though. They were seen as competitors in a competition for both trainers and participants.
But the training of trainers' courses are ours, since we know that we do not pay horrendously, it is a chance for us in the competition with the other organizations, to convince people to work for us. ( The relationship to functionally similar organizations can be characterized as non-hierarchical and indirect. Relationships to other functionally similar organizations which are non-hierarchical and direct include cases of cooperation between two organizations. Such cooperation refers to both publicly funded adult education organizations and other adult education organizations. With the colleagues here, we meet if we have topics to discuss. For instance, in February we met to prepare a public discussion with the candidates for the election. It was about education policy. These things we do together. (Interview Feldberg, Mr. Müller) And then there are the other leaders of the organizations here in the region. We have a lot of cooperation, a lot of coordination as well and we have a regional panel, the round table of adult education [...]. (Interview Feldberg, Mr. Müller)

Discussion of findings and conclusion
In this paper we introduced neo-institutionalism into the theoretical approach of political economy of adult learning systems. This approach has allowed us to take the organizational level into account as well as to specify the environment of adult education organizations by focusing on the actors of the systems Desjardins distinguishes as well as on the constellation of relevant environmental actors.
Applying this theoretical perspective, we have re-analyzed data from the GLOW study which aimed at identifying new governance regimes in publicly funded adult education organizations. The research question for the study asked leaders and program planners of public adult education organizations which actors they relate to as important in their environment, and how they describe the interdependent relationship with these actors.
In line with our theoretical background we assumed that interdependent relations can take various forms, e.g. direct or indirect, hierarchical or non-hierarchical, and can exist between functionally different or similar organizations. The findings show that adult education organizations relate to both functionally different as well as similar organizations in their environment. If interdependency relationships are described between functionally similar organizations, they are non-hierarchical but never hierarchical (as we have assumed). They can be realized in cooperation as direct or in competition as indirect. In this case of publicly funded adult education organizations, hierarchical relations only extend to functionally different organizations in the environment, which usually belong to the broader economic, social and cultural system. Relations to functionally different organizations can be characterized both as hierarchical and non-hierarchical, with forms of coordinated action usually limited to direct interaction in hierarchical settings. From a theoretical background, we did not expect the few references to indirect forms of coordination of actions with relevant environmental actors. This last finding also needs to be explained with reference to the interview set-up in the GLOW study. It might indicate a methodical limitation since indirect relations are based on reciprocal observation, which is difficult to access in interviews.
However, we have made clear that this analysis was meant to explore-for the first time-the amendment to the original theoretical perspective. The results of the re-analysis provide first indications that the institutional conditions that shape adult learning systems can be specified with the help of neo-institutionalist ideas, particularly with regard to the adult education organizations involved and their institutionalized interdependent relationships with other (organized) environmental actors.
For farther-reaching insights, a more differentiated data basis is needed. As such, commercial organizations have to be added to the sample as well as non-profit organizations since those are embedded in different contexts and thus faced with different institutional expectations (and environmental actors). What is more, following on from the explorative phase, a standardized survey is to be carried out to broaden the claim of validity of the analysis. And finally, the analysis can be extended into an internationally comparative study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.