The Professor Will See You Now: Gnomes
4 min read
In an occasional series, Professor Philip Cowley offers a political science lesson for The House’s readers. This week: gnomes
How does your garden grow?
One of the earliest conversations I ever had with an MP was about how, when canvassing, he claimed to be able to predict which party someone would support based on the state of their lawn. At the time I thought this seemed unlikely, but in the decades since I’ve had multiple chats with MPs and activists along similar lines. Garden gnomes equal Conservative, apparently.
If you knew someone’s class, age and education, for example, how likely would you be to be able to predict their vote?
It’s not inherently implausible. It might be a pretty noisy cue, with lots of exceptions, and maybe one that would vary geographically, but there is plenty of research that shows political tribes demonstrate differences in other non-political aspects of their lives – Conservatives have tidier bedrooms, for example – so why not horticulture?
A fascinating research paper published last year tried something similar, albeit based on more conventional cues. If you knew someone’s class, age and education, for example, how likely would you be to be able to predict their vote?
The researchers tested this by showing respondents profiles of voters and then asking them to say how they thought that person had voted – either in the 2017 election or in the Brexit referendum a year earlier.
You try it. Imagine a 27-year-old white man with no religious affiliation; he is a graduate owner-occupier living in the East Midlands earning somewhere between £60,000 and £999,999; he considers himself working-class, as he did when growing up. How did he vote in 2017?
A neat aspect of this paper is that each of the profiles that respondents were shown were of real voters – those who had previously taken part in the British Election Study. As a result, we know exactly how they had behaved at the ballot box and can compare respondents’ guesses with the reality. (That 27-year-old voted Conservative.)
The paper concluded that, on average, voters did have an understanding of the factors that underpinned people’s votes. They were, for example, more likely to predict a high probability of a Leave vote when the individual really was a Leave voter; ditto for Conservatives. Plus, respondents who reported paying higher levels of interest in politics were more likely to get it right. The differences were not great, but they were statistically significant.
There is, however, a lag in how voters see the world. People tended to over-emphasise class differences – now less significant than they used to be – but underestimate more recent electoral cleavages such as age and education. Indeed, education was the only variable where people got the direction wrong – thinking voters with degrees were more likely to be Conservative than Labour, when the opposite was true.
One other aspect of the work really struck me. Respondents were asked to say how certain they were about their estimates – and they often went for high levels of certainty, saying that there was zero per cent or 100 per cent probability that someone had voted in a certain way, even though a moment’s thought would indicate that such a level of certainty was unwise.
In part, this just reflects how people struggle with probabilities – but it is also, I think, because we often talk about demographic characteristics as somehow determining someone’s vote, when they do not. They are probabilistic, not deterministic. Think, for example, of how we talked about how graduates voted in the Brexit referendum, even though there were millions of graduates who voted Leave. The same, no doubt, would be true of garden gnomes.
For understandable methodological reasons, the paper only tested binaries – Remain/Leave or Lab/Con. Yet even in the referendum, there was a third option (to not vote), and other elections have long since ceased to offer binary choices. Actually, predicting someone’s vote – based on their background or their lawn – is even harder than it used to be.
Your further reading for this week: N Titelman and B Lauderdale, Can citizens guess how other citizens voted based on demographic characteristics? Political Science Research and Methods (2023); D R Carney et al, The Secret Lives of Liberals and Conservatives: Personality Profiles, Interaction Styles, and the Things They Leave Behind, Political Psychology (2008)
PoliticsHome Newsletters
Get the inside track on what MPs and Peers are talking about. Sign up to The House's morning email for the latest insight and reaction from Parliamentarians, policy-makers and organisations.
