Identifying causal links between net zero policy and behaviour
Project reference: LE72
Application deadline: 10 April 2023
The drive towards Net Zero requires large scale societal changes in how we produce and consume energy. As well as directly investing itself, the government must also set the landscape of incentives and opportunities available to companies and individual citizens through taxation and subsidies to motivate the take up of more energy efficient technologies, investment in energy-saving infrastructure and behavioural adaptations such as changes in transport choice. Although economic theory can provide some guidance to these incentives, it is important to validate their effectiveness through rigorous analysis of data.
This requires determining reliable causal connections between government actions and subsequent changes in the choices and behaviour of individuals and companies, to determine which policies are working and which need to be adapted or abandoned. Although there are good data available on many key metrics, such as the take up of house insulation or installation of solar power generation, establishing these causal links is potentially difficult because of social contagion effects. Behaviour can be driven not only by the external incentives an individual is presented with, but also by the behaviour of others, such as neighbours, friends, or competing companies. As such, adoption of a new behaviour can depend critically on the structure of these social connections, and stochastically on chance occurrences such as adoption by highly influential individuals.
This project will focus on the statistical challenges of inferring robust causal links between external policies and subsequent adoption of new behaviours in a socially-connected community. The research will require mathematical modelling of complex social contagions, generation of observable statistics from this process, and the development of statistical analysis tools to identify causal relationships when aspects of the social network or contagion process are unknown. The research will make use of open data of key metrics (e.g. from data.gov.uk, Swiss Federal Office for Energy, the UK Power Network Open Data Portal), to inform modelling and test methodologies. A candidate will have a strong mathematical background (e.g. an undergraduate degree in mathematics, physics, engineering or computer science) and will also have experience in statistics and probability.