Our #DataImpactFellows programme (currently accepting applications until 26th March) has helped support early career researchers in different ways to develop data impact. Oliver Exton discusses his experiences teaching undergraduates how to use UK Data Service data.
Data impact can take many forms, from addressing research questions, to disseminating policy relevant conclusions, or employing data as a tool to enhance the teaching environment.
As a Teaching Fellow at the Faculty of Economics and Jesus College at The University of Cambridge, I have had the opportunity to supervise undergraduate students undertaking empirical dissertations using UK microdata. This experience has highlighted how high-quality data can enhance opportunities for undergraduate students to develop their research skills which can be employed in further study and future careers.
What is microdata?
Microdata is information at the level of individuals or firms which can be collected from surveys (such as the Labour Force Survey), censuses (the UK census) or administrative tax records (such as the Business Structure Database). This data contains information on many characteristics of the individuals or firms, and often provides information over time. The many micro datasets provided by the UK Data Service offer fantastic opportunities for undergraduate students to apply and develop their data analysis skills when conducting their own research projects.
Empirical research projects encourage students to address academic and policy relevant questions, whilst working through the whole research process including identifying a research question, finding data and developing an empirical strategy.
Finding appropriate data to be able to answer research questions can be challenging. Fortunately, the wide range of data available through the UK Data Service can be applied to answer research questions across a wide range of disciplines!
For economists, the longitudinal datasets available through the UK Data Service, including the Annual Survey of Hours and Earnings and the Business Structure Database, are particularly valuable as they facilitate research that can track individual workers and firms over time whilst also accounting for a rich set of observable characteristics.
Connecting undergraduates with microdata
Undergraduates using microdata for their research projects are able to apply their theoretical knowledge and develop their empirical data analysis techniques.
The opportunity to implement techniques learnt in lectures in a real-world setting is an important step for bringing data analysis to life! Some of the best ways to understand how different data analysis techniques, in the case of economics the different econometric techniques, are to implement them on data to answer research questions.
Many data analysis skills that can only be learnt when you get your hands dirty in empirical research.
Data analysis is not always straightforward, and a research project develops an understanding of the process of managing and organising data. Datasets do not always come in a format that is immediately ready for data analysis and can require processing or merging with other datasets before the main analysis can be conducted. All of these skills can then be employed in future research and are transferable to many careers where data analysis is becoming increasingly important.
The opportunity to supervise undergraduate research projects has also been an invaluable development experience for me as an early career researcher.
I have developed management skills through helping students to identify research interests, develop project plans and provide support to complete their projects. Using secure microdata also brings additional challenges in ensuring data access and working with multiple stakeholders to set this up – a process which was helped by the profile of my being a Data Impact Fellow. Finally, it has been incredible rewarding to help inspire the next generation of researchers to consider further study and how they can use data and quantitative skills in their careers.
Research I supervised
Max Bowling studied how UK firms have responded to the increased tariff uncertainty following the result of the Brexit referendum using the Business Structure Database from the UK Data Service and tariff data from the World Bank.
He exploited the substantial variation in the potential tariff rates that would be imposed if the UK were to drop out of the EU with no trade deal, to examine the firm response to tariff uncertainty on both the extensive margin of trade and the intensive margin of firm performance. He found that employment is 0.5 per cent lower in industries facing a one percentage-point higher potential tariff.
“It was a privilege to be able to use such a comprehensive dataset for my undergraduate dissertation. It gave me a real flavour of how an empirical research project is conducted, providing me with valuable experience and skills for research going forward. In particular, I improved my ability to manage and clean huge datasets using STATA to get them into a tractable form to run regressions on.”
Lawrence O’Brien investigated how import competition affects UK workers’ earnings and employment. Using the Annual Survey of Hours and Earnings from the UK Data Service and trade data from UN COMTRADE, he exploited the increase in Chinese import competition from 2000 to 2011 as an exogenous shock that causes a reduction in demand for workers in some tradable industries.
He found that workers who are more exposed to the shock have significantly lower cumulative earnings in their initial industry and lower wages outside this industry. Unexpectedly, managers have the highest earnings losses from the shock as they are caused to work significantly fewer years outside manufacturing.
“I really enjoyed the opportunity to work on microdata for my undergraduate dissertation. Working with such a detailed and broad dataset meant I improved my data analysis skills hugely and I could address a more ambitious and rewarding research project. The experience was one of the key factors motivating me towards further study.”
Alex Ball tested the predictions of heterogeneous firm models in international trade through an empirical investigation using the stock market reaction to Brexit using data from Bloomberg ad Bureau van Dijk.
Using the Brexit referendum result as an exogenous shock, and estimating the cumulative abnormal returns for a large sample of UK companies, he found that larger firms in terms of employment, turnover and market cap were less affected by the announcement of the Brexit referendum.
“The dissertation work was incredibly useful for career following university as it was the first time I had to work with a substantial data set. It also taught me how to think about ideas in a more critical manner and formulate my own ideas into a testable hypothesis.”
Oliver Exton @oliver_exton is one of our UK Data Service Data Impact Fellows. Oliver is an ESRC funded PhD student in the Faculty of Economics at the University of Cambridge whose research focuses on firms and workers in international trade.