We have announced the following dates for our upcoming ESCoE Economic Measurement Webinars. Please note that all webinars run 11.30am-12.30pm, and that some descriptions and titles are still to be confirmed. Look out for further details on our events page escoe.ac.uk/events.
14 January: Martin Weale (King’s College London and ESCoE)
‘Household Cost Indices‘
This talk explores how far it is possible to provide a theoretical framework for the Household Cost Indices. Four features are identified which distinguish the index from conventional consumer price indices: i) the index is calculated giving equal weight to each household’s expenditure pattern (democratic weights); ii) insurance premia are treated gross rather net of claims; iii) interest payments are included as a cost and iv) goods and services are accounted for when they are paid for rather than when they are consumed. Points i) and ii) are strongly supported. It is suggested that for theoretical coherence iii) needs to be expanded to include interest receipts as well as payments while noting that with democratic weights this may have little practical effect. Point iv) raises a number of questions. A coherent framework representing the life-time cost of consumption correctly would need to include payments made ahead of future consumption (saving) as well as payments made ex post (repayment of debt). At present the only expenditure item subject to the principles of iv) is higher education; the student loan scheme has many of the characteristics of a tax and the treatment in the household cost indices can be defended on those grounds. ONS intends to produce a variant of the index which reflects the capital costs of housing and some thoughts are offered on measurement of these.
28 January: Giordano Mion (University of Sussex and ESCoE) and Manuel Tong (NIESR and ESCoE)
‘The Impact of Offshore Profit Shifting on the Measurement of GDP: The Case of the UK‘
In this talk we analyse the global distribution of profits declared by multinational enterprises (MNEs) operating in the UK using the Orbis database. Our investigations cover the period 2007-2017 and focus on entities reporting non-consolidated accounts and belonging to corporate Global Ultimate Owners active worldwide. Our analyses suggest that, compared to actual declared profits, profits distributed according to a simple apportionment rule based on companies’ revenue shares within each MNE group would look quite different. In particular, MNEs operating in the UK reported in 2017 41 billion GBP (representing about 1.9% of UK GDP) more than what they would have reported based on our apportionment rule. In this light, the UK was in 2017 a net winner in terms of global MNEs’ profit shifting. The situation was actually reversed back in 2007, with MNEs operating in the UK reporting less profits than those arising from our apportionment rule. A closer inspection of the whole period 2007-2017 reveals a smooth change with the UK moving from a loser to a winner position mainly through changes in declared profits of UK-owned MNEs. We subsequently extend the analysis by examining industry-specific patterns and conduct a number of robustness checks concerning the apportionment rule and the companies involved in the analysis while pointing to a number of limitations of our approach related to difficulties arising in dealing with Crown Dependencies, Branches, Special Purpose Entities and Family Trusts.
11 February: Ivan Petrella (University of Warwick and ESCoE)
‘Modelling and Forecasting Macroeconomic Downside Risk‘
We document a substantial increase in downside risk to US economic growth over the last 30 years. By modelling secular trends and cyclical changes of the predictive density of GDP growth, we recover an accelerating decline in the skewness of the conditional distributions, with significant, procyclical variations. Decreasing trend-skewness, turning negative in the aftermath of the Great Recession, is associated with the long-run growth slowdown starting in the early 2000s. Negatively skewed predictive densities arise ahead of and during recessions and are often anticipated by deteriorating financial conditions. Positively skewed distributions characterise expansions. The model delivers competitive out-of-sample (point, density and tail) forecasts, improving upon standard benchmarks.
25 February: Andrew Clark (Paris School of Economics)
‘COVID-19, Lockdowns and Well-Being: Evidence from Google Trends‘
The COVID-19 pandemic and government intervention such as lockdowns may severely affect people’s mental health. While lockdowns can help to contain the spread of the virus, they may result in substantial damage to population well-being. We use Google Trends data to test whether COVID-19 and the associated lockdowns implemented in Europe and America led to changes in well-being related topic search-terms. Using difference-in-differences and a regression discontinuity design, we find a substantial increase in the search intensity for boredom in Europe and the US. We also found a significant increase in searches for loneliness, worry and sadness, while searches for stress, suicide and divorce on the contrary fell. Our results suggest that people’s mental health may have been severely affected by the pandemic and lockdown.
11 March: Thies Lindenthal (University of Cambridge)
‘Towards Accountability in Machine Learning Applications: A System-testing Approach‘
A rapidly expanding universe of technology-focused startups is trying to change and improve the way real estate markets operate. The undisputed predictive power of machine learning (ML) models often plays a crucial role in the ‘disruption’ of traditional processes. However, an accountability gap prevails: How do the models arrive at their predictions? Do they do what we hope they do – or are corners cut?
Training ML models is a software development process at heart. We follow best practices from software engineering and define a system testing framework to verify that the trained ML models perform as intended. Illustratively, we augment two real-estate related image classifiers with a system testing stage based on local interpretable model-agnostic explanation (LIME) techniques. Analysing the classifications sheds light on some of the factors that determine the behaviour of the systems. We show that cross-validation is simply not good enough when operating in regulated environments.
25 March: James Mitchell (Federal Reserve Bank of Cleveland and ESCoE)
‘Censored Density Forecasts: Production and Evaluation‘
This talk develops methods for the production and evaluation of censored density forecasts. Censored density forecasts quantify forecast risks in a middle region of the density covering a specified probability, but ignore the magnitude but not the frequency of outlying observations. A new estimator is proposed that fits a potentially skewed and fat tailed density to the inner observations, acknowledging that the outlying observations may be drawn from a different but unknown distribution. An application using historical forecast errors from the Federal Reserve Board and the Monetary Policy Committee at the Bank of England illustrates the utility of censored density forecasts when quantifying forecast risks after shocks such as the global financial crisis and the COVID-19 pandemic.