We are curious about how growing institutional complexity affects societal transformation processes such as sustainability or digital transformation. In our article, published in the Journal of Management Studies, we shed light on how the emergence of new institutional logics – defined as key societal patterns – and increasing conflicts with existing logics disturb and change the German energy market. By using a mixed-methods machine learning approach, we show how the German energy field is becoming divided into two increasingly separate fields due to an accelerated increase in complexity – for which we are introducing the novel concept of constellation complexity. This is a particularly important finding for all transformation researchers because it suggests that conflicting logics and complexity can both slow down and accelerate transformation processes.

Why is growing institutional complexity central to better understand transformation processes?

Scholars argue that societies and markets are shaped by different institutional logics that lead to very different assumptions, values, and ultimately actions and routines in these fields. Just think of the logic of the market and the growth paradigm and, conversely, a sustainability logic that focuses on circularity and balancing the different interests of a large number of stakeholders. Different societies or markets have specific or adapted logics, such as the big data logic in the context of AI and large language models (LLM). Such logics can be in conflict with each other and thus trigger changes in companies and societies, for example when the big data logic collides with the idea of resource-saving processes, which is related to the sustainability logic.

Now we argue that it makes a significant difference whether there are two logics competing in a field or whether there are four, five or in our case even seven logics competing for dominance. From this we develop a model of growing complexity, which distinguishes between the processes of increasing and accelerating complexity. While increasing complexity has been extensively studied in literature and practice, processes of accelerating complexity have received little attention, but have great potential to better explain radical change of companies and fundamental societal transformation processes.

Studying growing complexity with a mixed-methods machine learning approach

As institutional complexity is a multi-layered phenomenon that changes over time, it is important to study a wide range of time authentic and longitudinal data. We propose to combine for instance publicly available texts such as media data (e.g., press and/or social media) with data from organizations (e.g., annual reports) and legislative texts. To manage and meaningfully evaluate the high amount of data, we specifically use a mixed-methods research design that combines topic modeling and qualitative content analysis.

Why accelerating complexity can both slow down and accelerate transformation processes?

In our article, we show how complexity escalates and how this acceleration in complexity leads to a field or a market starting to diverge and separate into different segments. In addition to the scientific relevance of these findings, we also discuss a number of practical implications. Most importantly, our study shows that accelerating complexity can be a serious barrier for the diffusion of a sustainability orientation (sustainability logic) in the entire energy market. Although one part of the field has focused on sustainability issues (e.g., renewable energies), the other largely separated traditional part is still deeply embedded in rather unsustainable practices (e.g., coal power generation).

Additionally, due to company demergers, the traditional parts of the companies have had to engage little with sustainability questions, which impede the dissemination of more sustainable practices. However, whether this leads to a greater retention of unsustainable practices in the long run and slows the sustainability transition process is a question that merits more attention. A possible counterargument is that the emergence of a sustainability-dominated segment may facilitate the development of new sustainability-oriented companies in a protected space. Other research has shown that, in unsustainable settings, even sustainability-oriented organizations are often not able to work sustainably, because of long-established institutional constraints. Thus, separation as a consequence of an accelerating complexity is an important phenomenon to better understand fundamental transformation processes.