Not one Brexit: How local context and social processes influence policy analysis

Abstract

This paper develops an empirical agent-based model to assess the impacts of Brexit on Scottish cattle farms. We first identify several trends and processes among Scottish cattle farms that were ongoing before Brexit: the lack of succession, the rise of leisure farming, the trend to diversify and industrialise, and, finally, the phenomenon of the “disappearing middle”, characterised by the decline of medium-sized farms and the polarization of farm sizes. We then study the potential impact of Brexit amid the local context and those ongoing social processes. We find that the impact of Brexit is indeed subject to pre-Brexit conditions. For example, whether industrialization is present locally can significantly alter the impact of Brexit. The impact of Brexit also varies by location: we find a clear divide between constituencies in the north (highland and islands), the middle (the central belt) and the south. Finally, we argue that policy analysis of Brexit should consider the heterogeneous social context and the complex social processes under which Brexit occurs. Rather than fitting the world into simple system models and ignoring the evidence when it does not fit, we need to develop policy analysis frameworks that can incorporate real world complexities, so that we can assess the impacts of major events and policy changes in a more meaningful way.

Agent-based modelling for Urban Analytics: State of the art and challenges

“…The group also develops spatial ABMs and microsimulations on long temporal scales (10-50 years). These models allow us to address different types of research questions such as the impact of a socio-economic shocks on an urban area and its neighbourhoods [78,79], or the impact of Brexit and changing international relations on food security of cities [80,81]. These larger-scale transitions and the mid-to long-term ramifications can be hard to pick up in fine-scale data of everyday activities, and require different types of data such as regional or national demographic and economic data, as well as data on migration and social networks.…”

“…over small spatial areas, short time periods, or restricted to a narrow set of behaviours-but recent work is delivering models that can operate in much larger contexts. They are being used to simulate systems as diverse as transport [34], crime [17,24] and farming [80,81], and work towards 'digital twins' at the Turing aspires to include ABMs alongside other modules that are already working together [10]. In the long run, efforts to better track uncertainty in ABMs [46], coupled with closer collaboration with policy makers through the Turing, will begin to foster the use of ABMs as important policy-making tools.…”

Can Social Norms Explain Long-Term Trends in Alcohol Use? Insights from Inverse Generative Social Science

“…The number of candidate models is typically small and constructed by hand. For example, in a recent POM study, Ge and colleagues enumerated 16 different model configurations that exhaustively tested combinations of including or excluding four different mechanisms driving trends in Scottish farming (Ge et al 2018). In its standard form, POM does not consider any procedure for refinement or evolution of model structures; however Cottineau and colleagues introduced an incremental approach to ABM structural development, known as evaluation-based incremental modeling method (EBIMM), where mechanisms are gradually added and refined by the modeler in order to generate the target emergent phenomenon (Cottineau et al 2015).…”

Evaluating the implications of Brexit for research collaboration and policy: a network analysis and simulation study

“…Since then there has been a dramatic rise in the number of publications on Brexit and its potential impact on the UK’s performance across various sectors 1–3. A number of quantitative studies have been produced on the impact that Brexit might have on the UK’s economy, trade and industries 4–8. However, the majority of these studies in the scientific and healthcare literature have resulted in editorials and letters expressing individuals’ as well as scientific and industrial bodies’ concerns 1 9–13.…”