This blog is hosted on Ideas on EuropeIdeas on Europe Avatar

Making use of #opendata: EP Intergroups and the network of UK MEPs [2x updated]

Two weeks ago, I have blogged about European Parliament intergroups – 28 groups on various subjects which EP members can join to support a cause or to focus on special issues.

As the data is not easily accessible and usable, I have produce an open data set with a list of all EP intergroups and their members. And I have promised to take a look into the data with the help of network analysis.

Here some first results (click on the image for full size):

A quick explanation of what you see:

  1. Every node is one MEP, 74 in total. As the UK only has 73 MEPs, this data set will include at least one MEP who has already left the Parliament during this term.
  2. Every link represents at least one joint intergroup; thicker links represent more joint intergroups. Nodes that are not linked are not member in any intergroup.
  3. Node colour represents the political groups: European Conservatives and Reformists (black), Socialists & Democrats (red), Liberals (yellow), Greens (green), European United Left (pink), European Freedom & Democracy (grey), non-affiliated (white)
  4. Node shape represents gender and node size represent number of intergroup memberships (max. 8 )

What you can see in the network picture above is that UK MEPs’ membership in intergroups is pretty much shaped by their left-right political group affiliation. The Conservatives, EFD & non-affiliated members cluster together as do the Socialists, Labour, Greens and the United Left.

[Update:] For Social Network Analysis specialists I’ve uploaded a spreadsheet with a number of centrality measures for the nodes in the network depicted above.

The picture becomes even more clear when we only keep links for UK MEPs who are members in at least two joint groups (39 in total; non-linked nodes erased):

Here you can see that the picture one can already discover in the full network is even more striking in the reduced network.

From such a picture one could start looking into specific sub-clusters like the three ECR women on the left or the green-liberal sub-network on the upper left. Through which groups are these MEPs linked, do they work together beyond the membership that links them because they share common interests?

Or one can look into indivdual MEPs like the liberal Bill Newton Dunn on the lower right who seems to be more connected to the Conservative cluster. And indeed, looking into his Wikipedia CV linked above, one can see that formerly he actually was a member of the ECR group.

These are the kind of insights and stories network analysis can provide and this is why open data is quite important for political science, journalism and a better understanding of politics, including in the European Union.

If I have time, I’ll play around with the data a little more. Maybe you have particular ideas or questions?

Update 2: In the comments, there was the question whether one could not include the names of the intergroups in the picture. It is possible, as you can see below, although the picture gets a little more confusing (click on the image for full size).

Images produced with visone. Data transformation with the help of iGraph for R.

10 Responses to Making use of #opendata: EP Intergroups and the network of UK MEPs [2x updated]

  1. Just to put you right on a couple of genders – Ashley Fox is a man not a woman; Marta Andreasen is a woman not a man. Apart from that, good stuff.

  2. avatar Ronny Patz says:

    @ Nicholas: Thanks for pointing this out to me. I used an existing dataset for information on MEPs affiliation, gender, and nationality, so I never checked this myself. Will correct it in my original data set.

  3. avatar Jack Blumenau says:

    This is really great. I was wondering if it was possible somehow to indicate the subject matter of the shared inter groups as this would may mirror to some extent the traditional party programmes/areas of interest.

  4. avatar Ronny Patz says:

    @jack It is possible, for example by showing a network map that includes both, the intergroups and the MEPs. Visually, it get’s a little more confusing, but I’ve updated the post with an according picture.

  5. avatar Yan Dupas says:

    Thanks a lot for this interesting way to look at intergroups. Since I’ve worked on intergroups myself, what is striking me is the fact that indeed intergroup membership seems to go along party lines, while those informal meetings of MEPs pretend to be cross-party and presents this characteristic as an asset to pre-cook European legislation. So my question will be (but I guess you will need more than network analysis to answer): what can you deduce from this new finding about the role of intergroups in European policy-making?
    Unfortunately I have no clue how network analysis works, but could you do the same ranking MEPs according to their role in intergroups (i.e. President or Vice-President) or according to their actual attendance rate (I know some intergroups like the one on animal welfare publish this kind of data)?
    Thanks a lot in advance and good luck for the rest!

  6. avatar Ronny Patz says:

    Hi Yan,

    one things needs to be said, namely that the data used here is just about UK MEPs. I’ve been playing around with the full data set, and I found that party polarisation seems less obvious with all 740 MEPs included than in the UK sample. As I found the full data set to be too messy for a (comparatively) quick blog post, I was just trying to see whether the UK sample would provide me with a result that was a little easier to interpret.

    When it comes to trying to analyse also the role of MEPs based on functions or actual participation, one would indeed need to check every individual website of those intergroups that actually have a website EP website on intergroups only provides the chairpersons) and provide this information.

    But you are right, in order to make sense of intergroups, it is not enough to do network analysis based on membership – even if Jon Worth in his follow-up on my post made use of this network map to tell a story about an individual MEP. The advantage of the data set used here is that it is – supposedly – complete, given that we have the lists for all intergroups. I’m suppose that the study of participation or group leadership would be less systemic.

    The main reason why I was doing network analysis is that this is what I currently do for my own research (not on intergroups), and I want to see whether the methods I use are applicable and relevant for other circumstances than the ones that I focus on professionally. It’s also experimenting with the software I use.

    I’ll see how I will continue blogging on this subject. It’s kind of interesting but I’m not sure what would be the most fruitful path to follow. I’d be glad if some more people involved in/with intergroups would offer their perspective(s) on how one could make sense of (the system of) intergroups. But maybe it’s worth looking which of the groups actually proved participation lists and to see whether actual participation tells a different or the same story than the one that can be told about the UK MEPs data set.

  7. avatar Bruno says:

    Hi there,

    Great blog post, thank you. I coordinate the Intergroup on LGBT Rights in the Parliament. The PDF files you find on the EP website are updated only once a year, when we’re required to declare our membership and financial interests (if any) to the administration.

    Get in touch if you’re interested in more research on Intergroup membership, I’d be interested in helping out if I can.

  8. avatar Ronny Patz says:

    Hi Bruno,

    thanks for the information. I’ll see whether I’ll do more research on intergroups, this effort was more a proof of concept that social sciences in the 21st century can take a call from an academic paper (see previous post on which this one is based), gather certain types of data quite quickly, make it accessible to the public and then start discussions whether and in how far one can draw conclusions from that or not.

    See for example Jon Worth’s follow-up post and the discussions in the comments there:

    The usual academic process would have taken a year before this would have gone live and discussions might have started, and I think we can be better than that today. 🙂

  9. Pingback: The structure of EU politics: Story-telling about European Parliament committees | Polscieu

  10. avatar Nils Ringe (University of Wisconsin-Madison) says:

    My co-author Jen Victor (George Mason University) and I have a book in production with the University of Michigan Press titled “Bridging the Information Gap: Legislative Member Organizations as Social Networks in the United States and the European Union,” which offers a comparative view of what we call legislative member organizations, or LMOs, such as caucuses or intergroups. Aside from offering a broad overview of where such groups exist, and what predicts their presence or absence, the book is based on a nuanced comparison of caucuses in the US Congress and intergroups in the EP. The book uses a mixed-methods approach (quantitative, qualitative, and network methods). Expected publication is spring 2013.

UACES and Ideas on Europe do not take responsibility for opinions expressed in articles on blogs hosted on Ideas on Europe. All opinions are those of the contributing authors.