A better description of the features and functionalities of ICAs over time and space, and of the transfer of functions from one archetype to another, would allow us to understand this “phenomenon” better, also in a present-day context. What kind of functions do these institutions have in society and how do state and market interact with them? In their own form, they often also had a major impact on the functionalities state and market adopted over time. This is for example visible in the impact of labour unions on the introduction of laws for better working conditions, but also in the adoption of insurance schemes in 19th-century factories, following mutualist insurances set-up by workers themselves. Identifying the evolution of functionalities of ICAs over time, can as such also improve our understanding of changes on the market and state-level. As has been demonstrated in for example Germany, the formation of many new but relatively small (in number of members) energy cooperatives had a massive impact on the supply of renewable energy. Understanding the appearance and role of ICAs in society, could be of great value for society at large, with so many major societal challenges ahead.

The differences in functionalities between various archetypes, and why some types emerge, become popular but then often also disappear again, should however not only be described but also explained. The long-term comparison of archetypes shows two major elements that may contribute to the answer that come down to matters of size and heterogeneity in terms of resources and members. First, ICAs had a different take on resource heterogeneity over time. Whereas in the early modern period, archetypes such as guilds managed to accumulate many different purposes such as the setting of minimum quality and price, collectively buying raw materials but also providing insurance, their 19th- and 20th-century counterparts often focussed on a specific purpose, such as insurance. Secondly, early modern ICAs also followed another scaling strategy to deal with increased membership from those that emerged in the 19th century. Early modern ICAs, such as commons and guilds preferred to limit their number of members and did so by for example splitting their organisation in smaller units when needed. The Guild of Saint-Eligius in Amsterdam (set-up in the 14th century), for example, originally comprised both smiths and carpenters, but soon the latter formed their own guild. The ship carpenters remained within the Saint Eligius-guild until 1517, but then joined the guild of the Amsterdam boatmen sailing abroad (buitenlandsvaardersgilde). Shortly after the Reformation of Amsterdam, they finally formed their own specific guild of ship carpenters and mast makers. In the 20th century we see the opposite development of ICAs: instead of splitting and specialising, small cooperatives within the same sector increasingly merged into larger companies, but subsequently struggled increasingly with sticking to their cooperative principles due to their very large number of members. This was for example the case with the merger from the Boerenleenbanken to the Rabobank in the Netherlands, although the original Boerenleenbanken kept their autonomy as individual, local cooperatives. Recently, the local cooperatives were dissolved with one single, very large cooperative as a result.

Thus, although their specialisation strategy seemed to work, the first wave ICAs were very vulnerable to top-down dissolution measures by governments as they lacked supra-local organisation. Hence the quick and overall disappearance by – at the latest – the middle of the 19th century of most commons and guilds across Europe. ICAs of the second wave could more easily strengthen their movement by creating substantial critical mass both in membership and resources (e.g. Rabobank, FrieslandCampina). The current wave of ICAs, in turn, seems to react against the decrease in member-involvement the latter strategy brought along, by limiting the size within ICAs and instead connecting individual ICAs on a sectorial basis in a network (e.g. Nederlandzorgtvoorelkaar.nl). This present-day solution differs from both earlier scaling-strategies.

From these differences in ICAs over time, it becomes clear that size and heterogeneity in terms of resources and members must have played an important role in terms of diversification between archetypes but also in their survival strategies over time. Strangely enough, these aspects have received lesser attention so far among scholars of specific ICAs, such as commons and cooperatives. The focus has primarily been on how the institutional design can encourage cooperation and prevent free riding within ICAs. As such, our knowledge on the impact of changes in size and heterogeneity is in fact still limited. Concerning the members, it has been well-described that self-governing institutions rely heavily on group norms and reciprocal behaviour between their members, as becomes clear in e.g. studies on cooperatives. Game theory and principal-agent theory both emphasise that cooperating for socially optimal outcomes, and enforcing the mutually agreed rules, is actually very difficult to achieve. Research on the production of collective goods or the governance of common-pool resources has shown that specific social factors will promote cooperation, yet sufficient levels of cohesion, group identification, and internal communication must be maintained at all times. Many studies suggest that when the number of members is large, a participant may assume that her/his own actions will have little effect on the outcome. Scholars have been trying to determine whether small or large groups are more likely to cooperate successfully within an ICA-context and what this threshold level would be. Smaller groups are often presumed to allow for greater interaction and social cohesion, which ensures cooperation and avoids freeriding. Generally, it is assumed that as group size increases, a) the need for leadership becomes more manifest to manage operations, b) the amount of resources available for problem-solving may increase but due to increasing diversity within the group, cooperation and consensus-making may become more difficult, and c) the tendency to develop bureaucratic procedures increases. Increases in group size can reduce opportunities for frequent interaction, and thus fewer opportunities for reputation-building and mutual monitoring, both of which are thought to have a positive impact on cooperation. Most (experimental) studies on the effect of group size on cooperation however concentrate on the here and now and neglect changes in group size and composition over longer periods of time. Precisely because longevity is an objective of many ICAs, which may therefore be subject to significant change in contextual factors (e.g. wealth, occupation, dependency on resources, but also religion) that could shape individual members’ willingness to contribute to the collective effort, we need to have a closer look at changes in group size and heterogeneity over time. There are also good reasons to believe that what constitutes the group size or group composition varies depending on the ICAs development stage. Young ICAs may require a fundamentally different structure of members and resources than more established ICAs. A longitudinal/historical approach to how ICAs deal with variations in membership could greatly increase our understanding of the resilience of individual ICAs as well as of the movement as a whole.

As demonstrated by some of the larger European cooperatives today (most of which have emerged in 1880-1920), membership growth may conflict with the principles of member participation in decision-making processes. Zamagni underlines that today the ‘satisfactory administration of larger cooperatives is an unsolved issue everywhere’ and calls for more theoretical elaboration of cooperative managerial practices for large cooperatives.

The evolving economic and social contexts may also have a huge impact on how the group of members is composed and how the members perceive usefulness of the collective resources. The PI’s commons’ research has for example shown that if commoners, e.g. due to impoverishment, are no longer capable of keeping cattle, the right to pasture cattle on the common becomes less important, and may keep commoners from actively contributing, or even encourage them to freeride on other resources. This example demonstrates the importance of including also the resource heterogeneity, and not the size/amount, in our study. We can assume that higher resource heterogeneity would allow for pooling resources within an ICA and this could improve the individual members’ economic position, improve the individual utility of participation and thus form an incentive to cooperate. As such, resource heterogeneity could advance successful collective action. At the same time, however, multipurpose organizations, which offer a wider variety of resources, are institutionally more complex. Constant monitoring and good leadership are required to avoid overuse. The work by Poteete and Ostrom demonstrates that large and/or heterogeneous groups in natural resource-ICAs are however not necessarily a problem and that specific institutional settings can actually moderate size. Specific changes to the institutional design (such as punishing freeriding, enhancing overall and within-group enforcement, improving social capital across groups and among group members) can indeed have a positive effect on cooperation within groups and also moderate increasing group heterogeneity, but it remains unclear whether the impact of changes in group size and heterogeneity can also be moderated effectively over longer periods of time.