The E3 modelling of 4CMR began with a distinctly macroeconomic view, dividing the economy into a variety of sectors of economic activity. We refer to this modelling as "top-down" because it uses information on the total economic activity, total energy consumption and/or total carbon emissions from a sector. The result is a model of the aggregate behaviour of a sector, without proposing that this behaviour is representative of any specific individual, organisation or community in that sector. There is then an implicit assumption in simulating the effects of E3 policies on this system that the internal characteristics of any one sector remain the same over time, even if the total amount of economic activity, energy and/or carbon emissions of a sector change over time.
The advantage of such top-down models is that they don't require detailed understanding of these internal characteristics, such as the division of activities between different companies or communities within a sector. They don't require complex theories as to how individuals take decisions within those sectors. The focus is always on the aggregate behaviour of a sector when various kinds of "shocks" such as rising energy prices or regulatory requirements are applied. This allows the model to use historical, macroeconomic data to estimate the parameters, such as price elasticity for energy use in each sector. It provides a powerful empirical base to the modelling.
A weakness, however, is that such top-down models don't provide insights into how shocks and policies might influence the behaviour of individual actors within a sector. They provide insights into the aggregate influence of these shocks and policies so long as the mix of actors and their decision rules remain roughly the same over time, but don't provide much help in understanding how policies and incentives might target, and hence influence the different individual actors whose actions ultimately make up the aggregate behaviour studied by the top-down models.
Enter the bottom-up models. These models reflect the fact that in important ways, the behaviour of an entire sector is the aggregate of behaviours by individuals and organisations. Bottom-up models describe how these individuals and organisations take decisions that influence energy and carbon. They allow the simulation of policies that might influence different actors differently within a sector. They allow use of data on how people conduct economic activities, how they choose to build buildings and infrastructure, how they choose to live within these, how they choose to move about. We call these models "agent-based" because they treat aggregate behaviour in a sector as arising from the individual decisions of many actors, each behaving according to relatively simple rules. Think of the "Mexican Wave" at a football match, which looks quite complicated but is actually each individual behaving by the simple rule that if the person to the right stands up, I stand up; if that person sits down, I sit down.
The weakness of these bottom-up models is that one must assume that the aggregate behaviour in an economy is simply the sum of the behaviours of the actors, and they require information on the relative numbers of actors that have the different characteristics that lie at the heart of decisions.
Through our staff in 4CMR, and through our many partners in the Department of Land Economy and elsewhere at the University, we have expertise in and experience of these individual decisions in a variety of sectors - energy systems, buildings, infrastructure, transport, industry. Hence we are constructing bottom-up models that capture these decisions and allow us to understand how the policies, planning practices, finance innovations, etc examined at the macroeconomic level might also be examined at the level of individual agents.
To get the best of both worlds of modelling, we are active in developing, and then joining up, both top-down and bottom-up models so we can explore not only how policies drive aggregate economic, energy and carbon behaviour in a sector, but how these same policies can be tailored to change the ways in which individual actors take decisions that drive the economy. Our work focuses especially on understanding the ways in which national, bilateral and multilateral negotiations on climate policy might proceed even in the absence of a global policy framework, and how companies in the supply chains of major corporations can be provided incentives to decarbonise their contributions to Scope 3 emissions.
The first product of this research is an ABM built within NetLogo, which you can download here. It was generated for the paper "An agent-based model of global carbon mitigation through bilateral negotiation under economic constraints" by Crawford-Brown, Liu and Silva, submitted recently. Watch this space for the paper as it moves through review and publication.