So, it being the 10th anniversary of that Gordon Brown - Mrs Duffy moment, there's a lot of people on my TL claiming "Labour would have won but for that moment" & similar, displaying two of my favourite political biases: partisan bias and agency bias. 1/?
2/? For starters: No, Mrs Duffy didn't matter. The whole incident had no discernable impact on ratings of Brown, ratings of Labour, voting intention or anything else we have measures for. There's nothing there. No effect. Nothing. Nowt. Nada. Labour even won back Mrs Duffy's seat
3/? So why, then, is the Mrs Duffy moment treated as so totemic. There are, I think, (at least) two bad explanations for this and (at least) one good one.
4/? The first bad expl is agency bias - people strongly favour agency explananations ("x happened because Mr Y did/said this to Mrs Z") over structural explanations ("x happened because of this matrix of complex societal factors")
5/? Partly, this is just because structural explanations are more complicated. But I suspect we are also just wired to prefer explanations involving actors and motives - anthropomorphism is a big thing. This is a political variant.
6/? The second bad expl is partisan bias, which is indeed the dominant one on display today: "If it wasn't for those nasty people in the right wing press blowing up something meaningless, the good leader of my good tribe would have led a good government and been loved by all" etc
7/? This of course interacts with agency bias, something we see particularly in the borderline conspiracism strong partisans use to explain defeat: "we only lost because the usual powerful group of shadowy people from [bad tribe] conspired to mislead everybody" etc etc
8/? So, people like tribe based explanations and they like explanations which are stories involving people doing stuff. Even better, stories which combine both tribes and individual incidents. This leads to a lot of the worst nonsense written on what matters in elections
9/? Indeed, the tendency is so pervasive and persistent that academics founded a book series giving a comprehensive analysis of each British General Election in order to combat it. I'm currently working on the latest entry, another source of my interest in today's flurry of myth
10/? What about the good explanation? Well, IMO, the Brown/Duffy incident didn't *change minds* but it did *reinforce existing beliefs* by *symbolising* various complicated aspects of voters' views of Labour at the tail end of the New Labour govt
11/? In particular, Brown's switch between polite discussion of Mrs Duffy in public and angry dismissal of her in private reinforced longstanding concerns voters had that New Labour (& politicians in general) were untrustworthy.
12/? The conversation was also on immigration, the most neuralgic of issues for Labour after 2004. Anti-imm voters saw the incident as the perfect confirmation that Labour dismissed their concerns as bigotry.
13/? Pro-imm voters saw the incident as a demonstration that voters *were* bigoted, and the subsequent circus and Brown "teagate" kowtowing apology as evidence that the media were indulging and reinforcing said bigotry, and Labour were too quick to do likewise.
14/? So in a sense the incident's power, such as it was, reflects another powerful bias: confirmation bias. A campaign stumble turned into a Rorsasch test in which everyone saw what they already believed, reflected back to them.
15/? Labour would still have lost the election if Mrs Duffy had never popped out to get milk. But, after the fact, Mrs Duffy provided lots of people with a symbol around which they built (often contradictory) stories of Labour and Brown's failings. That's how she matters. /ends/
