On Sunday, July 20th, I took part in the panel on Sub-national Parliaments at the 23rd World Congress of Political Science. The Congress is organized by the International Political Science Association, and is one of the few forums that brings together political scientists from around the world to exchange ideas and get up to speed on the latest research.
I presented on my dissertation research into all-party groups, which are informal, non-partisan bodies that members of a legislature form in order to mobilize around a particular issue or topic. The premise of the paper (click here to download) was to see whether the explanations that Susan Webb Hammond developed to explain participation in all-party groups at the US Congress can be extended to Westminster legislatures – specifically those of Canada, the UK, and Scotland.
Hammond found that participation in all-party groups was higher among those Members of Congress with lower margins of victory who were presumably looking for new ways to connect with their constituents. New Members of Congress also saw the groups as a way to gaining policy influence at an earlier stage in their career, and to build networks and demonstrate their leadership skills in the hopes of securing more influential positions in the legislature. While members of Westminster legislatures certainly share these concerns and goals, they also must contend with much higher levels of party discipline.
While the paper covers the three cases mentioned above, the presentation focused only on the Scottish Parliament. In some ways, Scotland is an ideal case since unlike the UK and Canada, the Scottish Parliament has had a system in place to regulate and monitor the operation of all-party groups ever since it was established (the Parliament was created in 1999). Thanks to the wonders of the Internet Archive, I have been able to obtain a copy of the registry of all-party groups for every session of the Parliament, making it possible to track how membership patterns changed over time as the Parliament evolved and Members of the Scottish Parliament (MSPs) responded to changes in their electoral fortunes.
While seeing if Hammond’s findings can be extended to all-party groups Canada and the UK was fairly straightforward, Scotland posed more of a challenge since it uses a Mixed-Member Plurality electoral system whereby 73 MSPs are elected from constituencies using single-member plurality, and 56 MSPs are elected by proportional representation at the regional level. As a result, the two types of MSPs had to be examined through separate regressions since there was no single metric for margin of victory that could be employed for both cases.
My largest finding was that the two different types of MSPs have different levels of involvement in all-party groups, with regional MSPs were much more likely to take part. However, there was limited replication of Hammond’s results. In particular, there was no relationship between years served and group involvement, and no link between marginality and participation among constituency MSPs. Ultimately, the only parallel was that participation was higher among regional members with lower margins of victory.
Notably, an unexpected finding was that participation in all-party groups was higher among women MSPs, although curiously the relationship was only significant among women MSPs elected at the constituency level. This results mirrors my findings from Canada and the UK, where women MPs also had higher rates of cross-party activity.