Seminars in 2017/18


Speaker: Gerhard Toews (Oxford)

Date: 28 March 2018

Title: What’s in a wedge? Misallocation and Taxation in the Oil Industry (with Radek Stefanski)

Abstract: Resource misallocation explains a large part of cross-country productivity differences. Measuring gaps in marginal products of labor and capital across countries and firms allows for a quantification of the extent of this misallocation. However, it is typically not informative on the source of the misallocation. We address this problem by using novel data from the oil industry to pin down both, the extent and the source of misallocation in the rest-of-the-world versus the United States. We confirm the existence of sizeable gaps in marginal products across production units, but show that these disappear once we account for direct taxation. This provides strong evidence that gaps in marginal products - and hence productivity - are largely driven by differences in tax policies rather than more indirect distortions.


Speaker: Dr Matt Dickson (University of Bath) [BS-HERU joint seminar]

Research interest: Public policy, health, education, applied micro-econometrics using large-scale administrative data

Venue: 25 April 2018

Title: The Causal Effects of Education on Adult Health, Mortality and Income: Evidence from Mendelian Randomization and the Raising of the School Leaving Age

Abstract: On average, people with more education are healthier, wealthier and have higher life expectancy than those with less education. In this paper, we investigate the causal effect of education on health and health behaviours exploiting two natural experiments. The first is a compulsory schooling reform in the UK in 1972 which generated a discontinuous increase in average education for the cohorts affected. The second uses perturbation of germline genetic variation associated with education which occurs at conception, known as Mendelian Randomization. The first strategy identifies a ‘local average treatment effect of education’ for those at the lower end of the distribution. In contrast, the increases in education associated with Mendelian Randomization occur across the education distribution. Much of the previous literature suggests that the differences in health outcomes associated with education are largely driven by selection into education on factors that also impact the outcomes. In contrast, the two independent sources of variation we exploit largely imply consistent causal effects of education. The congruence of the estimates from the two instrumental variables strategies strongly suggests that education causally affects health outcomes and that the size of the impact is similar across the education distribution.


Speaker: Professor Graeme Martin (University of Dundee)
Date: 23 May 2018
Research Field: Professions, healthcare and managerialism
Title: The changing work experience of hospital consultants in NHS Scotland: A case of ‘status quo bias’ rather than ‘enlightenment’?

Abstract: The increasing hybridization of healthcare organizations, combining often conflicting institutional logics, has led many doctors to experience severe professional role conflicts. Recent research has shown how doctors have been able to navigate a ‘middle way’ between traditional medical professionalism as an anchor point and the New Public Management and Political Democratic logics now seen to govern decision-making in the NHS.  This middle way has been most obvious in the creation of ‘hybrid medical leaders’, leading some researchers to hold out the prospect of doctors being able to ‘change their spots’ (Reay et al., 2017)

In this presentation, Graeme will present the findings from an extensive, mixed methods study of hospital consultants in NHS Scotland, conducted with colleagues from Dundee and Glasgow Universities.  The findings question this typically agentive and optimistic middle way drift in the literature, which foreshadows a hybrid recategorization of doctors and assumes an ‘ultimate goal’ is to engage doctors in management by having them ‘come over to the enlightened side’ (the strapline for the Academy of Medical Leadership and Management).  Instead Graeme and his colleagues identify: (1) a significant ‘status quo bias’ among the large majority of consultants in NHS Scotland, who see themselves as increasingly deprofessionalized, and (2) a distinct dividing line between ‘enlightened’ medical leaders, non-clinical managers and so-called jobbing consultants.  This dividing line is of major practical importance because of significant pressure on consultants to provide a consultant-led service, in which leadership is seen as an essential part of their role requirements.

The research, which was funded by the BMA is ongoing, and provides the underpinnings for an impact case to be submitted to REF2021.  The early findings were widely covered by the news media, TV and radio, and were presented in the Scottish Parliament.  They have also had an impact on policy and practice in NHS Scotland, and have led to invitations to do diagnostic and organizational development work with consultants and medical leaders in Sydney and a hospital trust in England.   Graeme will also focus on this impact and public engagement aspect of his research in his role as research director in UDSB.


Speaker: Dr Gregory Emvalomatis (University of Dundee)

Research interest: Environmental Economics, Industrial Organization, Production Economics, Efficiency and Productivity Measurement

Date: 30 May 2018

Title: Semi-Parametric Efficiency and Productivity Analysis using Gaussian Processes

Abstract: Since the formal definition of productive efficiency, parametric and non-parametric techniques for efficiency and productivity analysis developed in parallel. Over the last two decades advances in the non-parametric stream of the literature relaxed some of the restrictive assumptions, but very little has been done from the parametric side to bridge the gap. This paper proposes a fully Bayesian semi-parametric method for efficiency and productivity analysis based on Gaussian processes. The proposed technique frees the researcher from having to specify a functional form for the production frontier and is shown to outperform non-parametric methods in the presence of noise in the data and perform almost as good as correctly specified parametric models. The technique is applied to a panel dataset of US electric utilities, where Total Factor Productivity is estimated and decomposed using alternative methods.



Speaker: Prof Nils Braakmann (University of Newcastle)

Research interest: Microeconometrics using large-scale datasets, labour economics, health economics, economics of crime

Date: 6 June 2018

Title: The impact of student diversity on student outcomes at university and beyond – Evidence from English administrative data (with Stephen McDonald)

Abstract: We investigate the role of increased student diversity during university studies for degree and subsequent labour market outcomes. Using unique administrative data for all undergraduate degree students entering English universities between 2008 and 2010 and several identification strategies, we find that increased diversity is beneficial for students’ degree outcomes, while having peers with the same background also improves outcomes. These effects interact with each other, are quantitatively heterogeneous across institutions and student subgroups, but the effects of increased diversity are positive in the majority of cases. Employment outcomes unambiguously improve with exposure to diversity, independent of how this is measured.


Speaker: Professor Paul Thompson (University of Stirling)
Date: 27 June 2018
Research Field: Work, employment and society
Title: Connecting the workplace to the bigger picture some theory building reflections on the career of the disconnected capitalism thesis
Abstract: Theory building is the ultimate life blood of the social sciences. Yet the way that we produce and consume papers often means that we seldom think about the long-term conditions, constraints and choices that underpin it. This presentation offers a partial corrective through a reflection on the ‘career’ of a moderately successful concept I devised and developed – the disconnected capitalism thesis. In essence, this is an account of how financialization in the ‘bigger picture’ impacts (negatively) on workers and the workplace. The presentation sets out the origins, diffusion, departures, applications and impacts of the concept and makes some observations about theory building and ‘careers’.