Jevons Paradox, neither a paradox, nor a curse for new technologies

It has taken a few decades but many would agree that keeping global warming below +2C in 2100 is close to impossible and that adaptation is the major challenge in front of us. It is a distributed challenge (there is no single "us") because both initial conditions and warming will be very different from one country to another and from one social category to another . A global warming intensifies, adaptation will emerge as a necessity. For french readers, I recommend "Survivre à la chaleur" from 

Matthieu Glachant and Francois Leveque; for others, their argument in a nutshell is that adaptation will dominate the next three decades, because it can deliver short-term effects, it can be decided locally (does not require global perspective to be effective)  and because the pain-fear-action-reward cycle can be triggered individually (whether a person or a company).

System dynamics is a methodology for understanding, modeling, and analyzing complex systems over time using feedback loops, stocks, and flows. It helps simulate how different variables interact dynamically within a system, revealing patterns of behavior such as growth, decline, and oscillation. By mapping causal relationships and using mathematical models, system dynamics allows decision-makers to test different policies and predict long-term outcomes. It is widely used in fields like business, economics, healthcare, and environmental management to optimize strategies and improve system performance. For instance, CCEM is a System Dynamics model.

System Dynamics is a great framework to model, to assess and to simulate climate change adaptation, because of both the dynamic and systemic vision : each action has further consequences that change the problem. System Dynamics is well suited to describe situation with constant change, including the changes that are provoked by the system's actiors themselves. Climate change is indeed a complex problem that is evolving constantly as the world evolve. Any action taken has multiple impacts that may change the problem. For instance, decarbonation decisions may have unintended consequences (what CCEM shows very well) while adaptation decision may cause the carbon emissions to worsen (maladaptation ). 

The Jevons Paradox occurs when technological improvements that increase the efficiency of resource use lead to a higher overall consumption of that resource, rather than reducing it. Named after economist William Stanley Jevons, who observed this effect with coal consumption in 19th-century England, the paradox suggests that as efficiency lowers costs and increases productivity, demand rises, often offsetting the expected savings. This phenomenon is relevant in energy efficiency, where improvements in fuel economy or electricity use can lead to greater consumption due to lower costs and increased accessibility.

System Dynamics is well suited to modeling the introduction of a new technology that brings more efficiency to deliver an existing service, since we capture the set of relationships with the whole system. We can model the improved efficiency with lower cost, higher throughput, increased benefit of the associated service but also the raise of the matching negative externalities. The following figure shows a crude simplification where the value (in green) varies with the amount q of resources consumed (here showing a classical decreasing return). The resource quantity q is a function of time and can equally represent a yearly consumption or total consumption since the service started. The cost grows with the quantity q (red curve), and the chart represents with a dotted red line the introduction of a new tech that is more efficient. If the value formula does not change, the rebound effect is visible (yellow area on the chart) : cost reduction will be used to expand the scope of service delivery, but the value function is likely to change as well. In a world that is constrained by resources, value may go down for other reasons and the resulting equilibrium may be that technology improvement is used to sustain  usage, not to grow it.

What matters in a SD simulation is that the green value curve is not a static parameter, it results from the interaction between sub-components of the model. Hence as adaptation unfolds, the value profile changes, and the "value potential surface" may grow, stay stable or shrink (the yellow surface represents the variation of potential if efficiency grows and everytghing else stays the same.

A SD model will react differently depending on the other subsystems and the other links. It may show a rebound effect, not a paradox but a logical choice, in some cases and may, in other cases, show that new technology improvements do translate in overall resource consumption decrease. Another way to put it is that “Jevon paradox” is not a curse (it does not necessary occur) and when it occurs (such as with the introduction of the Watt steam engine) it is not a paradox from the viewpoint of the industry manager who decides to grow because of increased efficiency (reduction of costs for a unit of value extraction). 

Thinking about adaptation with a system dynamics framework is a great way to escape the mental model of the "rebound effect".