2.1 The Wall Street Crash of 1929 as a tipping point in an energy transition

Evidence that the Great Crash was a tipping point in an energy transition is provided, in part, by changes to the New York Stock Exchange (NYSE) indexes for coal and oil companies from 1926 to 1937 (Figure 1). The index of coal stock prices was relatively constant between 1926 and October 1929, before falling rapidly by approximately 90% by July 1932. Most notably, the coal index did not recover in the next 5 years. Behavior of the stock price index for oil production and refining companies was quite different; it followed the pattern of the aggregate stock market index, rising prior to the crash, sharply falling, and then recovering. By 1936, the stock price index for oil companies was back to its 1926 base value, but the price index for coal companies was still down by 85%.

Not all energy transitions necessarily involve major depressions (Kennedy, 2023b) and the circumstances of the current transition to renewables are different from those of 1929. Nonetheless, the case of the Great Crash demonstrates how a tipping point can be understood from a material and energy flow perspective, and how an energy transition entailed a fundamental transformation of a complex socio-economic system, including—amongst others—dominant practices of being mobile, as cars and airplanes became dominant mobility providers, replacing railways in the early decades of the 20th century.

2.2 Tipping points and socio-metabolic regimes in Paris

In the case of Paris, there are several long-term urban metabolism studies that couple metabolic flow evolution with a description of infrastructure development and changes in legislation, as well as the interplay between stakeholders, including government officials, companies, and citizens. These studies provide the necessary empirical insights to characterize social tipping points for the flow of energy and water.

For both these flows, one or more events triggered a transition process which eventually led to a socio-metabolic lock-in effect, and which, in turn, led to a tipping point. In the case of energy, the trigger was a shortage of wood fuel, which led to the adoption of coal through a transformation of supply areas and the utilization of newly created railroads (see Figure 2; Kim & Barles, 2012).

By developing long-term MFA studies, Barles and colleagues made possible the a posteriori identification of transitions between metabolic regimes. By complementing the MFA with institutional and infrastructural layers, it was possible to unveil the triggers and tipping points within these transition cycles. Finally, while these studies touched upon the average consumption per capita of energy and water flows over time, the relationship between infrastructural artifact and daily practices such as heating, hygiene, and nourishment, was less documented and could become a topic of further exploration.

2.3 Tipping points related to phosphorous flows in Switzerland

The case of phosphorous (P) over time illustrates how environmental tipping, health prevention, and metabolic information triggered legislative changes which in turn impacted the metabolic system. Table 1 shows, in a simplified way, how a combination of several triggers led to or induced socio-political tipping and changes in legislation. We will exemplify some of the dynamics in two main episodes.

Second, increasing phosphorus prices and supply uncertainty in the 2000s raised the question to what extent Switzerland could become more self-sufficient regarding its P supply. Indeed, according to Mehr et al. (2018), the total amount of P losses in waste management increased from 9000 t/year in 1989 to 13,300 t/year in 2015. Several studies (e.g., Binder et al., 2009; Binder & Jedelhauser, 2014; Lamprecht et al., 2011; Mehr et al., 2018) had identified the potential to increase circularity by extracting P from wastewater in the same order of magnitude as the current P imports via fertilizer. Based on these results and the supply uncertainty, the Ordinance on Avoidance and Disposal of Waste (VVEA) was introduced with the goal of “re-creating the P cycles between the waste sector and the agri-food system” (VVEA; Jedelhauser et al., 2018). Specifically, the VVEA asks for the technical recovery and recycling of P from municipal wastewater, sewage sludge, or sewage sludge ash and the re-utilization of P in meat and bone meal as of 2026. While it took almost 20 years from the first study to the mechanism's implementation, the final trigger was the perception of an acute P scarcity.

The example of P shows that crises related to environmental problems, human health (Bovine Spongiforme Enzephalopathie; BSE), or resource scarcity led to socio-political tipping and legislative reactions, and related changes in practices. For example, in the case of BSE, the use of meat and bone meal as animal feed for ruminants was banned, in 1990; specified risk material was fully removed from the feed chain in 1996; and animal byproducts were banned for fertilizer use in 2001 (FSVO, 2015; Mehr et al., 2018). In most cases (except BSE), MFAs or scientific metabolic studies created the basis for understanding the problems, developing solutions, and monitoring the outcomes. However, triggers were always needed to provoke a change in legislation and consequent effects on the metabolic system. Thus, metabolic studies can be essential for anticipating TPs and recasting the solutions for mitigating potential impacts.

3.1 Modeling tipping points in socio-ecological technical systems

To be able to capture TPs, models need to encapsulate nonlinearities and feedback mechanisms in systems over time that lead to fundamental structural changes in the modeled system. The mathematical and computational tools for establishing such models have advanced drastically in the last half-century (e.g., Fieguth, 2021; Gardiner, 2009; Strogatz, 2018; Thurner et al., 2018).

Techniques for establishing the appropriate form of a model in different application areas have also followed within the IE community. Particularly, current developments in LCA could provide relevant insights for analyzing the environmental impact of future trends in technologies. In addition, dynamic modeling of TPs could intersect with forward-looking work in LCA, whether using consequential techniques, considering future technologies or policies, tied to integrated assessment models (IAMs), or broadly using some form of transitional LCA (Mastrucci et al., 2024; Ventura, 2022).

Nevertheless, considering nonlinearities and feedback when modeling TPs in an industrial ecology context entails the following challenges:

Second, there is both a need and opportunity for advancing theory that establishes support for different model choices. The appropriate models for capturing the dynamics of TP in socio-ecological systems are likely to require the integration of theory from other disciplines, including social sciences, engineering, planetary sciences, and complexity science—see Thurner et al. (2018). Many industrial ecology professionals and researchers are likely versed in one or more other disciplines. This represents a strength in this community for creating the necessary collaborations to model the types of tipping points discussed herein.

Third, a key challenge is the need to include socio-economic dimensions (e.g., actors, institutions, decision trees, practices, and provisioning systems and their services; Schaffartzik et al., 2021; Rao et al., 2019). Models capable of representing possible tipping dynamics (toward, e.g., sufficiency; Gough, 2023) in socio-ecological systems need to incorporate socio-economic and ecological phenomena at a similar level of complexity (e.g., spatial or temporal resolution, differentiation of subsystems or processes and feedbacks). Models that represent social dynamics via agent-based modeling (ABM) and integrate that with biophysical phenomena (e.g., climate extremes, changes in land-use intensity and farming practices, and use of construction materials as well as their consequences) could be a promising approach here (Egger et al., 2022; Egger et al., 2023; Knoeri et al, 2014; Mayer et al., 2022).

Finally, given the data and theory challenges, it is likely that, for the study of TPs in the near term, it will be possible to show that multiple models fit similarly well despite each having different tipping point characteristics and susceptibilities. Subsequently, it may be more appropriate to use tipping point detection methods in concert with efforts to identify appropriate models. This suggests that industrial ecology methods such as MISO and EE-MRIO can be used complementary to TP research outputs and expanded with for example Integrated Assessment Models (IAM). Furthermore, they will be most useful in practice with methods for decision-making under uncertainty.

3.2 Early warning signals and establishing tipping point targets

The triggers and analyses presented suggest the question: what could be the early warning signs for future tipping points? Can we anticipate potential future tipping?

Early warning signals relate to different types of measures, indicating fast, frequent, and nonlinear changes with a high standard deviation in the variables measured (Scheffer et al., 2012). Further, the longer the time of recovery is following a perturbation, the more likely it is that the system is approaching a TP. This is supported in the literature on the modeling of nonlinear dynamic systems (Fieguth, 2021). Another early warning signal we could potentially identify is the delta of the growth rate. Exponential growth with a relatively short doubling time might indicate a rapid development, and thus, a rapid change of the system. Finally, frequent changes in events can be an early warning signal of an impending tipping point.

The potential for detecting TPs is also foundational to the concept of planetary boundaries (Steffen et al., 2015) of essential earth systems like the climate and ozone layer. Here, the boundaries are defined relative to zones of uncertainty around the tipping points in each system. The logic is that, for policy purposes, the planetary boundaries should be established conservatively outside the zone of uncertainty so as to reduce the risk of reaching one of the harmful earth-system tipping points. However, one needs to acknowledge that not all planetary boundaries are yet defined around tipping elements, for example, the land-use boundary is still rather arbitrary, demanding further research.

There are important links between planetary boundaries and industrial ecology TPS for sustainability. First, tipping socio-ecological systems toward sustainability is, in the aggregate, about keeping earth systems within planetary boundaries. Second, TPs for sustainability in socio-ecological systems should be established following a conservative approach concerning uncertainty similar to that used in the planetary boundaries approach. However, while the goal of the latter is to inhibit undesired tipping, that of socio-ecological sustainability targets is to ensure tipping.

4.1 Infrastructure, lock-ins, and social metabolism

In the future development of carbon-neutral technologies (e.g., e-mobility), the role and interaction between infrastructure development, legislation, and adoption at the individual level is crucial (Sharpe & Lenton, 2021; Kekäle and Helo, 2014). Recent cross-country analyses have shown that size and spatial patterns of built structures (buildings, roads, and railways) co-determine per capita levels of energy, GHG emissions (Haberl et al., 2023), and materials use (Duro et al., 2024) about as strongly as GDP. These analyses statistically established that the effect is additional to the well-known correlation between resource use and GDP, and that built structures are much more important determinants of per capita resource use than other factors widely studied, for example, population density, climate related heating energy demand, or fuel prices. IE is already well positioned to understand systemic metabolic impacts of infrastructure change such as how energy systems affect life cycles of other processes. With IE tools and, in collaboration with social scientists and economists, IE can contribute to the simulation of the dynamics of these processes, including the material perspective, and comprising the embodied emissions (Mohareb & Kennedy, 2014).

Furthermore, the high cost and long life of infrastructure provides lock-ins and constrains development pathways. Helmrich et al. (2023) state that infrastructure can be viewed as an “agent against transformation” given that we can identify self-reinforcing feedbacks perpetuated by “individuals, organizations and institutions.” This lock-in can be further reinforced when purely technological alternatives are considered (Helmrich et al., 2023; Markolf et al., 2018).

4.2 Cities and circularity

Given the key role of cities in climate transition, there is an urgent need to understand their dynamics from a coupled socio-ecological–technical perspective that accounts for nonlinearities (Bettencourt et al., 2007). With IE tools and linking up with social and political sciences, we could simulate the nonlinear dynamics of the urban environment, its potential development, and material consequences to finally derive information for knowledgeable decision-making (see Meirelles et al., 2021 for a first attempt). Indeed, TP has a role to play when simulating cities’ development as they “are largely built from the bottom up” and thus “if the essence or urban development is individual action, then a city can only be as smart as its citizens” (Batty, 2018, p. 177). Along these lines, during the last two decades, Batty and colleagues have studied the dynamics of urban territories using ABM and cellular automata (Batty, 2007, 2012) which could help modeling TPs at an urban level. More recently, these efforts were complemented with other methodological tools (scaling and power laws, Bettencourt, 2020, 2021; Roo & Rauws, 2012) to study the emergent properties of complex urban networks and build “an urban science.” Furthermore, technological development and emerging computational power has been used to develop urban observatories (https://newcastle.urbanobservatory.ac.uk/) and urban digital twins (Haken & Portugali, 2021) to measure, monitor, and simulate emergence and phase transitions in urban development.

Still, these analyses are relatively stock and flow agnostic and can also lack socio-political historical context which would help to explain how current practices were formed through metabolic lock-in effects resulting in a cascade of political, infrastructural, and other triggers. In the future, it would be relevant to cross-pollinate these two worlds by combining “urban science” methods with IE ones. For example, mapping the material stock and estimating material re-use in buildings is an increasingly active area of research and practice that could benefit from nonlinear dynamic stock models with economic, demographic, and technologic considerations. Further, the role of disruptions in causing changes in infrastructure patterns is an important topic requiring further consideration of potential long-term changes in urban metabolism (Magassy et al., 2023). Recently developed global maps of transport infrastructures (Wiedenhofer et al., 2024b) and buildings (Haberl et al., 2024) can provide valuable data for research in this field.

Another consideration for city researchers is to understand processes that lead cities to tipping points toward more sustainable directions. Examples include: London's implementation of congestion charging (Kennedy et al., 2005), Curitiba's revolutionary bus rapid transit system (Cervero, 1998) and rapid adoption of electric vehicles in Bergen, Oslo and other Norwegian cities (IEA, 2017).

4.3 Supply-chain disruptions

Given current geopolitical turmoil, an open question is how supply-chain disruptions related to conflicts, wars, chance events (e.g., blocked shipping routes), pandemics, or climate change related extreme events will affect future socio-metabolic patterns? Up to what point are resource systems resilient toward such events and when will they result in tipping points where new patterns emerge? How do they affect efforts to work toward more sustainable resource use? IE could contribute to addressing these questions by linking up with complex system modelers, social scientists, economists, and foresight researchers. Recently developed big-data approaches could also be instrumental here (Reisch et al., 2022).

5.1 Theory building

5.2 Closer collaboration with social sciences and economics

5.3 Identify and define TP indicators

5.4 Making use of new modeling technologies

5.5 Open science