Multi-agent, non-equilibrium, and network models of labour economics and financial contagion

The financial crisis of 2009 showed that the economy is not in equilibrium. That is, the economy is in a state where agents have incentives to modify their chosen actions. The recent COVID-19 pandemic has shown that the economy changes rapidly and that reaching an equilibrium can take years, if one is reached at all. However, assuming equilibrium has long formed the basis of many economic models. Although a non-equilibrium approach may seem better suited for answering economic policy questions during times of rapid change, previously, the lack of data and computing power made it difficult to build these type of models. Today, these factors are less constraining. This thesis contributes to the development of non-equilibrium economic models that can provide accurate answers to policy-relevant questions efficiently. The models we develop have the potential to enable fast economic analysis and can be run in near real-time. In the first part of this thesis, we focus on the labour market. We develop a model to analyse how workers move through an empirically derived occupational mobility network in response to labour demand shocks. We show that we can approximate the behaviour of the agent-based model with a discrete dynamical system. At the macro level, our model reproduces the Beveridge curve's dynamics, i.e. the negative relationship between vacancy and unemployment rate. At the occupation level, we study the potential impact of automation and the COVID-19 pandemic on employment. We find that the network structure plays an important role in determining unemployment levels, with occupations in particular areas of the network having few job transition opportunities. In the second part, we use multiplex networks to study global financial contagion. This work focuses on identifying systemically important countries, i.e. countries whose financial distress can propagate to other countries. We build a global financial multiplex network and develop a multiplex threshold model of financial contagion. We find that the number of systemically important countries can be twice as large when we consider a multiplex network structure compared to a monoplex structure. These two models have the potential to be used for policy design. Arguably, support and skill development programs should be directed towards workers who are likely to face longer periods of unemployment. Our model can help policymakers identify who these workers are. Our multiplex model of financial contagion represents a first step towards developing multiplex network models which policy institutions could use to prevent the spread of financial contagion.