The state and future of Political Methodologyon 11 January 2016
By Andreas Murr, Stephen Fisher and Paul Whiteley
A few weeks ago in the British Academy, the Quantitative Methods in Political Science specialist group of the Political Studies Association was launched with a one-day conference showcasing quantitative work in political science in the UK.
Qualitative and Quantitative
There have been various initiatives from various quarters focused on improving the quality of quantitative methods training at undergraduate and graduate levels in the UK, most notably the Q-Step programme. These have been across the social sciences and have often involved politics, government and international relations departments. While they are very much to be welcomed, they rarely if ever address issues with the development of qualitative and quantitative methodology for research in political science. Given that research on methodology is needed and that qualitative and quantitative methodology should not be divorced, it makes good sense to have a specialist group on political methodology in the UK to discuss and disseminate developments in this area.
Methods and Substance
It is great that the Political Studies Association are enthusiastic about having such a group as part of the PSA so that research on methodology can be better integrated with the study of substantive topics in politics. After all, in Britain there are many researchers with a substantive focus who are also seeking to make methodological developments for their fields and want to learn from others working on similar issues in other fields. The PSA Political Methodology group is not just for academics with a primary focus on methodology. The new group should help facilitate discussion between methodologists, as well as between methodologists and substantive researchers with an interest in methodology. These interactions we hope will help improve the quality of research and teaching, both in methodology and in substantive areas.
Maria Grasso, University of Sheffield
The Dangers of Hegemony
There is of course a danger that the methodology group becomes dominated by those from areas where quantitative methods are prevalent, such as electoral behaviour. While our launch event did reflect that bias, and also one towards quantitative methods, we very much hope that future events will succeed in drawing interest and participation from those in a broad set of substantive areas and from those advancing both qualitative and qualitative methodology.
The conference featured ten presentations and one keynote lecture. The topics ranged from the future of election surveys (Harold Clark) to how parties learn from successful parties in other countries (Hugh Ward) to methodological innovations for studying terrorism (Jeff Gill). Although the range of topics was broad, most presentations were focused on electoral behaviour, and so for the next events we would like to also feature presentations showcasing the excellent applied and methodological work done by UK based researchers in fields such as conflict studies, international relations, and political economy. With growing use of quantitative methods in international relations, conflict studies, security studies and in the text processing of documents and speeches it is clear that many sub-disciplines of political science can benefit from the collaboration between researchers who use these methods.
The same applies for the methods that the presentations used. They ranged from latent class measurement models (Anja Neundorf) to a hierarchical model cross-tabulated data (David Manley) and to text analysis with multiword expressions (Ken Benoit). Although the range of methods again was broad, all presentations were using quantitative methods, and so in the next conferences we would like to also feature presentations showcasing the excellent applied or methodological work done using qualitative or mixed approaches in the UK. It is clear that many sub-disciplines of political science can benefit from collaboration between researchers using different methodological approaches.
The State and Future of Political Methodology
On that note, we asked the eleven presenters what they think the state of political methodology is, what they think are the most exciting new developments, and why they think a political methodology group is necessary:
1) What do you think about the state of political methodology?
- Harold Clarke (University of Texas at Dallas): “Political methodology has made great strides over the past two decades. A new generation of political methodologists promises more to come. Political methodology is an intellectually engaging field of inquiry that attracts some of the very best young scholars.”
- Jeff Gill (Washington University in St. Louis): “Political methodology is easily the most dynamic, advanced, and visible subfield of political science. Recent advances in causal inference, Bayesian methods, hierarchical models, regression discontinuities, election forensics, panel data analysis, treatment of latent variables, and more are highly visible within the discipline and in other fields. Scholarship from political methodology is cited in medicine, sociology, psychology, economics, public health, statistics, and other fields.”
- Will Jennings (University of Southampton): “Political methodology by its nature is always in a perpetual state of innovation, but it is being pushed forward more than ever before by the greater availability of data from micro- to macro-level, advances in quantitative methods and training within the profession, and by the exponential growth of processing power that enables us to do more sophisticated things quicker. Perhaps more than any other subfield of political science, political methodology has to adapt to changes in our technological environment. These require us to constantly reflect on how we can marshal the methods available to us for making causal inference.”
- Anja Neundorf (University of Nottingham): “In general, quantitative political science is doing very well and might be dominating political science at the moment. Political methodology is key to this development.”
- Hugh Ward (University of Essex): “I don't think there is any shortage of activity among methodologists. We are importing sophisticated new methods all the time from geography, economics, sociology, and so on. The failing I see as an editor is more often to link the method used closely to a clearly developed theory. I think there is a danger we'll end up like (it was said) the U. of Rochester's PhD programme: all econometric wiz-kids but nothing to point the tools at. We need to integrate training in theorizing with training in methods. This is a general statement. Stirring things are happening in the UK to catch up, as a result of the efforts of lots of people. But we still lag a long way behind the US and, increasingly, Europe.”
Harold Clarke, University of Texas at Dallas
2) What do you think are the most exciting new developments?
- Harold Clarke (University of Texas at Dallas): “There are many exciting developments in political methodology. I will mention two that I find particularly interesting. One is the development of practical Bayesian methods and their application to a wide range of topics. A second is the revolution in survey research, especially growing recognition that high quality internet surveys can provide the large array of data needed to advance understanding of the dynamics of political attitudes and behaviour.”
- Will Jennings (University of Southampton): “Big data, while exciting, in itself is not enough for good political science. However it does provide opportunities for more thoughtful ways of measuring key political science concepts and phenomenon – it is now easier to extract large quantities of data on public opinion, elections, public policy, global conflict, etc. to use as the basis for quantitative analysis. It also is becoming increasingly possible to link data sources – and that is where the real scope for developments in answering important questions about politics lies.”
- Anja Neundorf (University of Nottingham): “Massive, new data availability in all subfields of political science. The use of online data. all kind of experiments (e.g. field, lab, survey).”
- Hugh Ward (University of Essex): “Hard to separate what I'm exited about from what is important, but I'd say new methods for estimating network effects ERGMs, latent space models and the like.”
3) Why do you think the new Political Methodology group is necessary?
- Harold Clarke (University of Texas at Dallas): “In the United States the Society for Political Methodology has played a key role in increasing the quality of research and graduate teaching in several subfields of Political Science and has done much to strengthen the reputation of Political Science as a serious scientific discipline worthy of extramural funding. A political methodology group in the Political Studies Association can play similarly beneficial roles in the UK.”
- Jeff Gill (Washington University in St. Louis): “A new political methodology group in the United Kingdom will provide a forum for the excellent quantitative scholars to communicate and collaborate. This is an important step towards building a community of empirical political scientists locally to contribute to methodological developments within the Political Science Association.”
- Will Jennings (University of Southampton): “The new PSA specialist group is really important in providing a vehicle for researchers who are interested in and excited by the application of quantitative methods to the study of politics. I hope it will provide for new conversations, both between those who specialise in political methodology and with other researchers who are interested in how these methods might apply to their own work but are not part of this community. The 2014 REF highlighted quantitative analysis as one of the areas of strength in political studies, and the group will provide a platform for showcasing developments in this field and for supporting the next generation of political methodologists. The specialist groups are the intellectual engines of the Political Studies Association and one of the things that we'll be hoping for will be joint projects between specialist groups so that quantitative methods are integrated across the whole political studies landscape - deepening and building upon our pluralist tradition to the study of politics.”
- Anja Neundorf (University of Nottingham): “Quantitative methods are used throughout different subfields of political science, from political behaviour to international relations. There is currently now group that represents this unifying dimension of our work, but we are rather split by substantive topic (e.g. EPOP).”
- Hugh Ward (University of Essex): “Going back to the state of the UK, we will only catch up by continuing to do the good work carried out by a lot of people to push things in their own departments. This isn't difficult in places like Essex where there is critical mass, but it's hard for a young quanto on her own in a typical UK department. We need group therapy! We also need an institutional voice in the PSA.”
Jeff Gill, Washington University in St. Louis
The presenters agree that political methodology is innovative by nature and relevant for advancing our understanding of substantive topics within and outside of political science. Many presenters highlight that technological advances such as internet surveys and increases in computing power are exciting new developments leading to innovations in political methodology such as practical Bayesian methods. Nevertheless, they also caution that technological advances such as big data cannot by themselves increase our understanding of politics without being used in a thoughtful way by a researchers with substantive knowledge. Overall, the presenters believe that the new Political Methodology Specialist Group can play a leading role in increasing the quality of political science research. The presenters point out that such a group will increase the collaboration and communication between researchers, leading to innovative research on important questions in political science. We are delighted that the launch event of the Political Methodology group in November was such a success in bringing people together, and we hope that you will consider joining the group or participating in its future events.
Andreas Murr is Lecturer in Quantitative Methods in Political Science at the Department of Politics and International Relations at the University of Oxford, Non-stipendiary Research Fellow at Nuffield College, and Non-stipendiary Lecturer at Lincoln College. He specialises in quantitative methods, particularly in Bayesian statistics and hierarchical models. His substantive research focuses on electoral behaviour, including models of decision making and election forecasting. He is co-editor of The Plot and treasurer of the Political Methodology Specialist Group of the Political Studies Association
Stephen Fisher is an associate professor in political sociology and fellow and tutor in politics at Trinity College at the University of Oxford. He is the author of various articles on elections, voting and public opinion. He is co-convenor of the Political Methodology Specialist Group of the Political Studies Association.
Paul Whiteley is Professor at the Department of Government at the University of Essex. His research interests involve examining the nature and significance of political participation, particularly electoral participation, and also in understanding the causes and effects of public opinion on politics. He is co-convenor of the Political Methodology Specialist Group of the Political Studies Association.