Integrated assessment modelling

• 1. Introduction

  ›

  Integrated assessment modelling sits at the intersection of climate science, economics, energy systems, land use, agriculture, health, and governance. A single question drives the whole enterprise: what will the future look like, and what will it cost to change course? These are not simple questions to answer, and the models built to address them are correspondingly ambitious. They do not predict the future. They sketch what possible scenarios look like, tracing pathways through a world shaped by human choices.

  At its heart, IAM tries to link the main features of society and economy with the biosphere and atmosphere inside one modelling framework. The goal is to support informed policy-making, most often in the context of climate change. But the models have also been applied to the Sustainable Development Goals, to patterns of conflict, to food security, and to development trends across Africa. The breadth of the enterprise is striking. So is the controversy it has generated.

  How reliable are these models? Who uses them, and what decisions rest on their outputs? And what happens when the models themselves become the subject of debate?

• 2. Building the Framework

  ›

  The word "integrated" in integrated assessment modelling carries real weight. These models span multiple academic disciplines: at minimum, economics and climate science. More comprehensive models also pull in energy systems, land-use change, agriculture, infrastructure, conflict, governance, technology, education, and health. The word "assessment" reflects the purpose: these tools exist to answer policy questions.

  There are two broad ways to classify IAMs. The first distinction separates models that quantify future developmental pathways and provide detailed sectoral information, called process-based models, from those that aggregate the total costs of climate change and climate change mitigation. The second distinction separates models that extrapolate verified patterns using econometric equations from those that determine globally optimal economic solutions from the perspective of a social planner, assuming partial or full economic equilibrium.

  Process-based models include notable frameworks such as IMAGE, MESSAGEix, AIM/GCE, GCAM, REMIND-MAgPIE, and WITCH-GLOBIOM. These have been the primary tools used by the Intergovernmental Panel on Climate Change to quantify mitigation scenarios. They have also been used to simulate the Shared Socioeconomic Pathways and to explore routes for staying within the 1.5 degree Celsius target agreed upon in the Paris Agreement.

  Not all process-based models assume rational agents or long-term market equilibrium. Non-equilibrium models, such as E3ME, which draws on econometric equations and evolutionary economics, and agent-based models like the DSK-model, take a different approach. They allow for systems that behave in messier, less predictable ways than the idealised social planner framework assumes.

• 3. The Price of Carbon

  ›

  Cost-benefit integrated assessment models serve a distinct purpose: calculating the social cost of carbon. That figure represents the marginal social cost of emitting one additional tonne of carbon dioxide into the atmosphere at any given moment in time.

  Three models in particular, DICE, PAGE, and FUND, have been used by the US Interagency Working Group for exactly this calculation. Their results have fed directly into regulatory impact analysis, meaning the outputs of these models have shaped real government decisions about emissions policy.

  The underlying logic is straightforward. Climate impacts are widely considered a negative externality, a cost that conventional markets fail to capture. If a government wants to correct that market failure, for instance by imposing a carbon tax, it needs a number: what does an extra tonne of carbon dioxide actually cost society? IAMs exist, in part, to provide that number.

  The difficulty is that the estimates produced by cost-benefit IAMs are highly uncertain, and researchers acknowledge they will remain so for the foreseeable future. That uncertainty has not stopped the models from being used; the effort to calculate the social cost of carbon is still considered valuable for understanding how specific processes affect climate impacts, and for illuminating one of the drivers behind international cooperation on climate agreements.

• 4. The Critics Speak

  ›

  In an October 2021 working paper, the economist Nicholas Stern argued that existing IAMs are inherently unable to capture the economic realities of the climate crisis under its current state of rapid progress. His critique was pointed. The models, he suggested, could not keep pace with the speed and scale of what was actually happening.

  Stern was not alone. A separate line of criticism holds that IAM-based analyses of climate policy create a perception of knowledge and precision that is illusory, and can fool policy-makers into thinking that the forecasts the models generate have some kind of scientific legitimacy. This is a striking charge: that the appearance of rigour can itself become a problem.

  Models that use optimisation methods have also drawn fire from researchers working in dynamical systems theory. That framework understands systems as changing with no deterministic pathway or end-state. The implication is a very large, or even infinite, number of possible future states, with aspects and dynamics that cannot be known from the current state of the system. Researchers describe this as "radical" or "fundamental" uncertainty. The charge is that optimisation-based IAMs suppress this uncertainty rather than confront it.

  In 2021, the integrated assessment modelling community itself examined gaps in what it called the "possibility space" and considered how those gaps might best be addressed. Some researchers have called for more work on alternative scenarios that have not yet received substantial attention, including post-growth scenarios.

• 5. Energy Blind Spots

  ›

  A specific technical complaint has also emerged alongside the broader methodological critiques. IAMs have been criticised for undervaluing and underestimating the role of primary renewable electricity, which is generated primarily through wind and solar power.

  Closely related concepts that modellers say receive too little attention include sector coupling, the electrification of end-uses, and power-to-X technologies. At the same time, critics argue that IAMs have overvalued the role of bioenergy and carbon capture and storage.

  Researchers point to a major discrepancy between the IAM modelling community and the energy system modelling community, two groups that operate with different tools, different assumptions, and, at times, different conclusions. The argument is that IAMs would be stronger if they drew more heavily on the findings of energy system modellers.

  That gap matters because the energy transition is central to nearly every decarbonisation scenario. If the models that underpin climate policy systematically misread the trajectory of wind, solar, and related technologies, the scenarios they generate may point policy-makers toward the wrong mix of solutions. Addressing that discrepancy is one of the concrete challenges the field has identified for itself going forward.