Approach

Our overall approach in this report is to synthesise knowledge about tipping points across multiple relevant disciplines spanning natural and social sciences. In general, we try to give primacy to empirical evidence of tipping point changes that have occurred, before considering potential ones that have yet to occur. In both cases, we try to provide underpinning theoretical evidence for tipping points. This means providing evidence of underlying causal mechanisms – notably self-propelling feedback mechanisms. This aims to counter the risks of promoting gratuitous alarmism (in the case of postulated negative tipping points) or naive optimism (in the case of postulated positive tipping points). 

Systemic risk

Risk is widely understood to be the combination of hazard (likelihood of an event), exposure (to impacts of that event), and vulnerability (of people/other species who are exposed to those impacts). This is the approach to risk used by the Intergovernmental Panel on Climate Change (IPCC). It can be applied to assess the risk of individual Earth system tipping points, as these can be imagined as isolated, specific events. But in reality they will not occur in isolation. As Sections 1 and 2 explore, they can interact with each other and with social systems, including having the potential to trigger negative social tipping points. As a consequence, a ‘static’ framing of risk that seeks to isolate the risk of specific events, soon runs into considerable difficulties when dealing with tipping points. As a result, we adopt a ‘dynamic’ framing of systemic risk (UNDRR, 2019). The key notion of systemic risk is that risk depends on how elements of affected systems interact with each other. We endeavour to highlight throughout the report what these interactions are and how they may affect risk.

Handling uncertainty

Tipping points are highly non-linear phenomena occurring in complex (and often adaptive) systems, where our knowledge of those systems is imperfect. The associated uncertainty may sometimes seem huge, and we must deal openly with it. The most fundamental uncertainty are unknown unknowns. It is quite conceivable that, when tipping events occur, they will happen in a way that we did not expect and may not fully understand. This report synthesises the known knowns and the known unknowns of tipping points, but recognises the existence of unknown unknowns and seeks to offer guidance that is robust to them.

For the known unknowns, uncertainty is present in both reducible and irreducible forms. Reducible uncertainty is that which arises due to a lack of knowledge. Throughout the report we highlight ways in which knowledge about tipping points can be further improved. Irreducible uncertainty is that which cannot be resolved just by learning or observing more. For example, tipping points can be triggered by random perturbations (‘noise’) that cannot be forecast in advance – such as the weather in the climate system, which is known to exhibit extraordinary sensitivity to initial conditions (chaotic behaviour). 

Despite the presence of irreducible uncertainties, it would be wrong to over-generalise that ‘all tipping points are inherently unpredictable’. There can still be predictive skill for some tipping points, it is just not a perfect predictive skill – as with the weather. Predictability exists because the systems we consider generally have a deterministic component to their dynamics – meaning they are governed by some laws that do not change over time. We may not know what those laws are, but we do not have to know them to detect their consequences. Notably, the phenomenon of critical slowing down gives measurable signals if and when a system is heading towards a tipping point. Usually we do know something about the laws governing the behaviour of a system, and sometimes we know enough to produce a process-based model of a system and its tipping point(s).  

We can usefully separate out some specific uncertainties surrounding tipping points, accepting the limitations (noted above) of a ‘static’ risk framework. 

First (and foremost) is uncertainty about whether a tipping point exists or not. We address that throughout the report, with reference to observations (past behaviour), theory (particularly regarding key feedback mechanisms) and models (including projections of future behaviour). For Earth system tipping points, we evaluate our confidence in their existence. We evaluate several candidates that we (currently) conclude are not tipping points, but nevertheless exhibit properties of non-linear change. These cases are clearly indicated. For tipping points in social systems, we evaluate their existence or not, but do not assign a confidence level to those assignments, because research is nascent in this area.

Second is uncertainty about how close (or far away) a tipping point is. Here ‘distance’ is best thought of in terms of some key driver (or drivers) forcing a system. An example is global temperature change in the case of climate tipping points. The uncertainty about the ‘location’ of a tipping point can be expressed in terms of an uncertain range in a key driver (or drivers). An example is the uncertainty in global warming at which a particular climate tipping point may occur. Within this uncertain distribution a most likely value may be assigned. This approach allows probabilities of a particular tipping point occurring under a particular forcing scenario to be derived and expressed in probabilistic (likelihood) language. While this is becoming possible for Earth system tipping points, it is not yet possible for social system tipping points. We discuss ways in which distance to a social tipping point could be derived, while recognising that, with multiple human agents continuously adapting their decisions and behaviour, that distance could be continually changing due to many drivers. 

Third is uncertainty about the consequences of crossing a particular tipping point. Evaluating this assumes a situation where the tipping point has happened. Hence the consequences can (in some cases) be more certain than the likelihood of the tipping point itself. They do, however, carry their own uncertainties. 

Fourth is uncertainty about who (or what) is exposed to those consequences. Evaluating human exposure requires a scenario or assumptions about the human population and its distribution, which carries its own uncertainties. These combine with the uncertainties in ‘mapping’ from consequences to those people. That ‘mapping’ may involve causal consequences propagating through complex networks.

Fifth is uncertainty about different people’s response to being exposed to the consequences. In the case of negative tipping points, this is termed vulnerability. In the case of positive tipping points, it can include being exposed to opportunities. In both cases responses depend on the state of individuals within families and communities, and on the state of wider social systems such as the global economy. 

Our normative position

The value judgements expressed in this report are based on applying principles of Earth system justice (Gupta et al., 2023). We all have a right to expect, and a responsibility to help secure, a world in which all people and all the other living things and ecosystems we depend on, can thrive in a way that does not diminish the ability of future generations to do and enjoy the same.

We have defined above how we assign ‘positive’ and ‘negative’ to particular tipping points, based on whether they are predominantly beneficial (positive tipping point) or detrimental (negative tipping point) for humans and the natural systems we depend upon. However, we acknowledge that one person’s positive outcome may be another’s negative outcome, and hence these assignments may be subject to debate. Here we expand on our rationale.

As a rule, the impacts of the Earth system tipping points are clearly ‘negative’ for most (if not all) people and many species. However, the actions driving us towards them may benefit some people in some ways – for example, through the extraction and use of fossil fuels. The impacts of smaller-scale social-ecological tipping points – such as abrupt collapse of fisheries or desertification – are also often clearly ‘negative’ for many participants in those systems. But again the actions driving the system past a tipping point may disproportionately benefit some people.

It is tempting to assign any and all actions – including social tipping points – that reduce the risk of negative Earth system tipping points as ‘positive’ – as they will reduce environmental harm for the majority, if not everyone. However, the associated social, technological and ecological changes can have costs as well as benefits that can be unequally distributed, calling for governance intervention. Otherwise, what is positive for a majority of people (or species) may still be deemed negative by some. 

Societies need to carefully consider the equity and justice implications of social tipping points that are ‘Earth system positive’, to try and minimise instances where they could be ‘socially negative’. This first means seeking to ensure they do not increase overall (global) harm and injustice, which means weighing up overall harms and benefits. Then, in cases where there are localised social injustices, good governance is needed to limit and mitigate these. For example, governments can provide social safety nets for those losing out – like supporting coal miners, their communities and regions in finding different employment and flourishing. At a deeper level, governance needs to decide the ‘welfare function’ – meaning what are we trying to maximise, what are we trying to minimise, and who do we accept is going to lose out. 

Governance

This brings us to our approach to governance of tipping points – whether ‘negative’ or ‘positive’. We take ‘governance’ to refer to rules, regulations, norms and institutions that structure and guide collective behaviour and actions, including the processes that create governance, which often involve politics, policymaking and mechanisms for holding actors accountable for their actions and omissions. We take a global governance approach that goes beyond state actors.

We consider not only governments as key governance actors and their intergovernmental initiatives, but also corporate and industry actors, civil society organisations, traditional authorities (e.g. village elders, monarchs), cities and municipalities, and transnational networks.

While attention to the threats posed by Earth system tipping points is growing, explicit governance efforts to address those threats do not yet exist. Section 3 addresses the key task of establishing a novel governance agenda and framework for Earth system tipping points, while recognising the difficulties for already-complex governance regimes to integrate a new set of challenges into their already-crowded agendas. Consequently, discussions about governing tipping points need to provide a clear and convincing logic for action, grounded in scientific knowledge, which this report aims to provide. 

The governance of positive tipping points poses its own challenges, which are addressed in Section 4. In particular, interventions designed for exponential and irreversible positive change can also carry the risk of exponential and irreversible negative change. A precautious, considered, systemic approach is therefore necessary to understand the potential consequences and to whom they might apply. Governance approaches that prioritise principles of equity and justice must anticipate and take steps to avoid risks and negative distributional impacts using compensatory and redistributive mechanisms. 

A particular risk is the creation of green sacrifice zones. These are ecologies, places and populations that will be severely affected by the sourcing, transportation, installation and operation of solutions for powering low-carbon transitions, as well as end-of-life treatment of related material waste (Zografos and Robbins, 2020). More broadly, we seek to avoid (and counter) climate colonialism, defined as “the deepening or expanding of domination of less powerful countries and peoples through initiatives that intensify foreign exploitation of poorer nations’ resources or undermine the sovereignty of native and Indigenous communities in the course of responding to the climate crisis” (Zografos and Robbins, 2020: p543). 
The desire to avoid damaging, potentially abrupt and/or irreversible Earth system and ecosystem tipping points is a key source of urgency in accelerating action on climate change and ecological crisis. Equally, triggering positive tipping points to accelerate action is a key response to that sense of urgency. However, for many Indigenous peoples and local communities who have faced the existential crisis of colonialism and who are now at the forefront of the climate crisis (Gilio-Whitaker, 2019), it may already be too late to avoid environmental injustices and so urgency to respond takes on a new perspective (Whyte 2021, 2020). Crucially, the urgency of tipping points needs to avoid overshadowing the slow process of rebuilding trust and relationships that have been broken through past harms (Whyte, 2020).