Regional Economic Forecasting in the United Kingdom in the Era of Big Data.

3-year PhD Scholarship, in collaboration with the Department of Economics.

  • Number of scholarships 1
  • Value Stipend: £14,652 pa. Fees: Home/RUK/EU Fee waiver
  • Opens 13 December 2016
  • Deadline 28 February 2017
  • Help with Tuition fees, Living costs
  • Duration 36 months

Eligibility

Candidates are required to have:

  • An excellent undergraduate degree with Honours in a relevant business, scientific/technological or social science subject
  • A Masters degree (or equivalent) will be strongly preferred
  • Students may also have other relevant experience or skills which are relevant to this project
  • Candidates who are not native English speakers will be required to provide evidence for their English skills (such as by IELTS or similar tests that are approved by UKVI, or a degree completed in an English speaking country).

Candidates should be available to take up study in the UK on 1st October 2017

*Whilst open to International candidates, please note that this scholarship covers Home/EU/RUK Fee rate only.

Project Details

Increasing regionalisation of policymaking in the UK makes having reliable and timely economic forecasts for its regions and nations essential. This need is especially acute in Scotland as a result of the ongoing devolution of fiscal powers to the Scottish Parliament, with the Scottish Fiscal Commission now charged with producing forecasts of the Scottish economy. The existing statistical models which are used at a national level, while able to combine data of different frequencies (M/Q/A) using mixed-frequency (MF) VAR models, fail to capture key dimensions of regional forecasting. The existing models do not, for example, constrain regional economic growth forecasts to ensure that -in aggregate- they reflect sensible national forecasts. Data availability and delays in the release of regional economic statistics pose further challenges for regional-level forecasting. Policymakers require two types of forecasts: very short term forecasts (‘nowcasts’) and longer-term forecasts. Empirical methods have been developed to tackle the challenge of nowcasting, and the ONS has funded the supervisors of this PhD to develop this kind of model for the UK regions as part of the Economic Statistics Centre of Excellence.

This fills a gap in providing a signal about the present rate of growth in the UK regions; aiding short term decision making. That still leaves the challenge of producing longer term forecasts of key dimensions of the UK regional economies, a topic on which there is little in the existing literature- particularly with respect to simultaneously modelling all the regional economies and ensuring that the regional forecasts constrain to aggregate national forecasts. This PhD will address this gap in the literature, and construct a model to produce simultaneous economic forecasts for all UK regions at different forecast horizons. The model will then be extended to incorporate advances that have been made elsewhere in developing methods for ‘Big Data’ forecasting models. The aim being to use our model with so-called ‘fat’ Big data, that is where we have a large number of candidate predictors (incorporating a large number of economic variables for each of the regions of the UK, and enabling each variable in each region to be a candidate predictor for every other region, one quickly gets into Big-Data VAR territory). Forecasting using models with these dimensions require the use of particular approaches e.g. LASSO methods to reduce the dimensions of the model. We therefore anticipate both applied and theoretical contributions from this PhD.

In summary, this PhD will develop a set of UK regional forecasting models and produce real-time forecasts of key macroeconomic variables. This involves: i) extending existing econometric methods to produce a MF model appropriate for UK regional forecasting, ii) implementing these models and evaluating their forecast performance, iii) extending the MF model to handle data of the dimensions envisaged in the Large VAR literature- this final element will encompass increasing the number of predictors but also increasing the range of variables for which are forecasts. This PhD represents a major step forward in the analytical forecasting capability available to UK regional policymakers.

How to apply

At this stage, we are inviting applicants to apply for the scholarship only. The successful candidate will then be asked to complete an application for PhD study at Strathclyde. 

All applications should include:

  • a cover letter indicating the candidate's relevant skills/experience and how they can contribute to this research
  • a CV and relevant qualification transcripts.
  • two references (please refer to guidance on references)

When sending the above documents please use the following file-naming convention: fullname_typeofdocument

For example,

Johnsmith_coverletter

Johnsmith_CV

Johnsmith_transcript1

Johnsmith_transcript2

Johnsmith_reference1

Johnsmith_reference2

Apply now by uploading your documents here. Please note that any incomplete applications or applications with files that do not follow the above format will not be considered.*

*We will keep your details on file to use when any other relevent scholarships arise.