Arthur Turrell

With Steve Rose and Mark Sherlock.
Large-angle Coulomb collisions allow for the exchange of a significant proportion of the energy of a particle in a single collision, but are not included in models of plasmas based on fluids, the Vlasov–Fokker–Planck equation, or currently available plasma Monte Carlo techniques. Their unique effects include the creation of fast ‘knock-on’ ions, which may be more likely to undergo certain reactions, and distortions to ion distribution functions relative to what is predicted by small-angle collision only theories. We present a computational method which uses Monte Carlo techniques to include the effects of large-angle Coulomb collisions in plasmas and which self-consistently evolves distribution functions according to the creation of knock-on ions of any generation. The method is used to demonstrate ion distribution function distortions in an inertial confinement fusion (ICF) relevant scenario of the slowing of fusion products.

With Steve Rose and Mark Sherlock.
High-intensity lasers can be used to generate shockwaves, which have found applications in nuclear fusion, proton imaging, cancer therapies and materials science. Collisionless electrostatic shocks are one type of shockwave widely studied for applications involving ion acceleration. Here we show a novel mechanism for collisionless electrostatic shocks to heat small amounts of solid density matter to temperatures of ∼keV in tens of femtoseconds. Unusually, electrons play no direct role in the heating and it is the ions that determine the heating rate. Ions are heated due to an interplay between the electric field of the shock, the local density increase during the passage of the shock and collisions between different species of ion. In simulations, these factors combine to produce rapid, localized heating of the lighter ion species. Although the heated volume is modest, this would be one of the fastest heating mechanisms discovered if demonstrated in the laboratory.

With Steve Rose and Mark Sherlock.
Particle-based simulations, such as in particle-in-cell (PIC) codes, are widely used in plasma physics research. The analysis of particle energy transfers, as described by the second moment of the Boltzmann equation, is often necessary within these simulations. We present computationally efficient, analytically derived equations for evaluating collisional energy transfer terms from simulations using discrete particles. The equations are expressed as a sum over the properties of the discrete particles.

With Karen Braun-Munzinger and Zijun Liu.
We construct an heterogeneous agent-based model of the corporate bond market and calibrate it against US data. The model includes the interactions between a market maker, three types of fund, and cash investors. In general, the sensitivity of the market maker to demand and the degree to which momentum traders are active strongly influence the over- and under-shooting of yields in response to shocks, while investor behaviour plays a comparatively smaller role. Using the model, we simulate experiments of relevance to two topical issues in this market. Firstly, we show that measures to reduce the speed with which investors can redeem investments can reduce the extent of yield dislocation. Secondly, we find the unexpected result that a larger fraction of funds using passive investment strategies increases the tail risk of large yield dislocations after shocks.

With Eleni Kalamara, Chris Redl, George Kapetanios, and Sujit Kapadia. Published at Journal of Applied Econometrics, but you can find an earlier working paper here.
This paper examines several ways to extract timely economic signals from newspaper text and shows that such information can materially improve forecasts of macroeconomic variables including GDP, inflation and unemployment. Our text is drawn from three popular UK newspapers that collectively represent UK newspaper readership in terms of political perspective and editorial style. Exploiting newspaper text can improve economic forecasts both unconditionally and when conditioning on other relevant information, but the performance of the latter varies according to the method used. Incorporating text into forecasts by combining counts of terms with supervised machine learning delivers the highest forecast improvements relative to existing text-based methods. These improvements are most pronounced during periods of economic stress when, arguably, forecasts matter most.

With Andy Haldane.
Macroeconomic modelling has been under intense scrutiny since the Great Financial Crisis, when serious shortcomings were exposed in the methodology used to understand the economy as a whole. Criticism has been levelled at the assumptions employed in the dominant models, particularly that economic agents are homogeneous and optimizing and that the economy is equilibrating. This paper seeks to explore an interdisciplinary approach to macroeconomic modelling, with techniques drawn from other (natural and social) sciences. Specifically, it discusses agent-based modelling, which is used across a wide range of disciplines, as an example of such a technique. Agent-based models are complementary to existing approaches and are suited to answering macroeconomic questions where complexity, heterogeneity, networks, and heuristics play an important role.

With Bradley Speigner, Jyl Djumalieva, David Copple, and James Thurgood.
Using a dataset of 15 million UK job adverts from a recruitment website, we construct new economic statistics measuring labour market demand. These data are ‘naturally occurring’, having originally been posted online by firms. They offer information on two dimensions of vacancies—region and occupation—that firm-based surveys do not usually, and cannot easily, collect. These data do not come with official classification labels so we develop an algorithm which maps the free form text of job descriptions into standard occupational classification codes. The created vacancy statistics give a granular picture of UK labour demand at the occupational level.

With Bradley Speigner, Jyl Djumalieva, David Copple, and James Thurgood.
What type of disaggregation should be used to analyse heterogeneous labour markets? How granular should that disaggregation be? Economic theory does not currently tell us; perhaps data can. Analyses typically split labour markets according to top-down classification schema such as sector or occupation. But these may be slow-moving or inaccurate relative to the structure of the labour market as perceived by firms and workers. Using a dataset of 15 million job adverts, we create an empirically driven, ‘bottom-up’ segmentation of the labour market. Our methodology demonstrates the potential of unsupervised machine learning in economics.

With Emma Duchini and Stefania Simion.
This paper studies firms and employees' responses to pay transparency requirements. Each year since 2018, more than 10,000 UK firms have been required to publicly disclose their gender pay gap and gender composition along the wage distribution. Theoretically, pay transparency represents an information shock that alters the bargaining power of male and female employees vis-à-vis the firm in opposite ways. As women are currently underpaid, this shock may improve their relative occupational and pay outcomes. Reputational concerns before employees, consumers and investors may also be key to understand firms’ response. We test these theoretical predictions using a difference-in-difference strategy that exploits variation in the UK mandate across firm size and time. This analysis delivers four main findings. First, pay transparency increases women’s probability of being hired in above-median-wage occupations by 5 percent compared to the pre-policy mean. Second, while this effect has not yet translated into a significant rise in women’s pay, the policy leads to a 2.8 percent decrease in men’s hourly pay, reducing the pre-policy gender pay gap by 15 percent. Third, combining the difference-in-difference strategy with a text analysis of job listings, we find suggestive evidence that treated firms adopt female-friendly hiring practices in ads for high-gender-pay-gap occupations. Fourth, a reputation motive seems to drive employers’ reaction, as firms publishing worse gender equality indicators score lower in YouGov Women’s Rankings. Moreover, publicly listed firms targeted by the policy experience a 35-basis-point average fall in stock prices in the days following the publication of their indicators.

With Ed Hill and Marco Bardoscia.
General equilibrium macroeconomic models are a core tool used by policymakers to understand a nation's economy. They represent the economy as a collection of forward-looking actors whose behaviours combine, possibly with stochastic effects, to determine global variables (such as prices) in a dynamic equilibrium. However, standard semi-analytical techniques for solving these models make it difficult to include the important effects of heterogeneous economic actors. The COVID-19 pandemic has further highlighted the importance of heterogeneity, for example in age and sector of employment, in macroeconomic outcomes and the need for models that can more easily incorporate it. We use techniques from reinforcement learning to solve such models incorporating heterogeneous agents in a way that is simple, extensible, and computationally efficient. We demonstrate the method's accuracy and stability on a toy problem for which there is a known analytical solution, its versatility by solving a general equilibrium problem that includes global stochasticity, and its flexibility by solving a combined macroeconomic and epidemiological model to explore the economic and health implications of a pandemic. The latter successfully captures plausible economic behaviours induced by differential health risks by age.

With Bradley Speigner, Jyl Djumalieva, David Copple, and James Thurgood.
Uncertainty still remains as to the cause of the UK's dramatic productivity puzzle that began during the Great Financial Crisis. Occupational mismatch has been implicated as driving up to two thirds of it. However, obtaining the high quality time series data for vacancies by job occupation that are required to measure occupational mismatch is a significant challenge. We confront this issue by using a weighted dataset of 15 million job adverts posted online that cover most of the post-crisis period and which enable us to test whether occupational mismatch still stands up as an explanation for the UK productivity puzzle. We find little evidence that it does, mainly because, relative to the data used in similar analysis by Patterson et al. (2016), our vacancy data imply greater heterogeneity in occupational matching frictions, a key determinant of the optimal distribution of labour across job types.

With Mirko Draca, Emma Duchini, Roland Rathelot, and Giulia Vattuone.
The pandemic was accompanied by a wave of adoption of remote work practices. This paper uses online job vacancy data to study how UK firms have adopted remote work. Overall, remote work increased by 300%. Our analysis finds little evidence that occupations have fundamentally changed to better accommodate remote work tasks, nor evidence of changes in the occupational composition of jobs. We find that the overall increase in remote working is driven by the increasing use of remote work at the firm level, especially among firms that were less likely to use remote work before the pandemic. This is consistent with changes in organisational practices or updated information about the viability of large-scale remote working.