Erica Johnson spoke in the second set of student presentations, and I chose to attend her talk (and not just because it was in the same room – I was very curious about her project on non-candidate blogging and elections). It turned out to be an interesting session for multiple reasons. Erica gave a very honest assessment of the difficulties of putting together a research project, and about half the room gave suggestions/comments/ideas; it was both reassuring and enticing at the same time. A point made on day one, about transforming your PhD topic from a noun into a verb, was recalled in the discussion.
Erica said: “My current research interest is too broad” – she’s certainly not the only one in that position! Someone (I think Ben Peters – Ben, correct me in comments if it wasn’t you) summarised the question that we should be asking – what keeps you up at night? (i.e. what are the big questions, what excites you, what are you worried about – all rolled into one).
Her own project, looking at non-candidate blogs during the 2000, 2004 and 2008 US presidential elections (from her beautiful vantage point in France), is the kind of thing that I love to read, although quite distant from my own research. In particular, she will be looking at three very different elections (one hasn’t even happened yet) with very different concepts of blogging and politics in each one.
Of course, there has been some work in this area already – and earlier in the day, we heard some of that work though Sunshine Hillygus of the school of politics at Harvard. Her research (and forthcoming book) deals with the impact of technology on campaigns and (in particular) how it affects candidates and their formation of policy. There were oodles of glorious statistics, with the general point being that despite the various ‘myths’ (that US voters are heavily polarised, that the floating voters are unmotivated, and that ‘divisive’ talk is just mobilising the base), microtargeting and the appeal to ‘cross-pressured’ voters (e.g. pro-life Democrats, anti-war Republicans, etc). The most interesting element, for me, was the discussion of the techniques of database-building and targeting. Instead of sampling and opinion polling alone, modern techniques often start with the (accessible and universal) electronic voter register, matched up with consumer data (bought), organisation membership lists (given) and so on, then some sampling and statistical model-building, and then heavily personalised contacts (I was impressed by the Democrats sending out flyers to pro-gun union members – “I care about two things – my guns and my job” – didn’t think they were that mercenary!). I doubt that it would be possible to be as thorough in the EU, due to data protection law, but certainly, it was part of the Blair revolution in some fashion). Hillygus wryly noted that all these special promises (and the almost exponential growth in the number of discrete policies in party platforms or debate speeches) lead to significant post-election problems…