At the latest local election in England, my school friend ran for office and failed to be elected. Because I've dabbled in marketing, he showed me the leaflet he posted through doors for feedback. I gave my specific opinion on his leaflet and he will take that advice on board for the next campaign.

Afterwards I realised that citing some literature on what aspects tend to be well liked would have been more useful.

Sample questions may be:

  • Does the number of pages in the leaflet affect vote share?
  • Does the leaflet being in black and white or colour affect share?
  • Does the percentage of words dedicated to policy affect share?

So my question is, are there any published studies of what aspects of campaign literature impact vote share in UK local elections?

  • 3
    FWIW, there is a fair amount literature out there that suggests that campaigning, at least when everyone acts on the assumption that it matters and makes a good faith effort to campaign, has virtually no effect on election outcomes in partisan races or in races with incumbents running. If I find some citations, I'll post an answer to that effect.
    – ohwilleke
    May 22, 2018 at 18:59
  • @ohwilleke - as per 538, political ads have meaningful impact when a candidate has low name recognition among electorate, and highly diminished impact once name recognition is higher. Not sure if that's universal or FPTP specific
    – user4012
    May 22, 2018 at 20:14
  • 1
    @user4012 Because of multiple parties and block voting, a lot of local election campaigning in England and Wales targets party recognition, rather than candidate recognition. On occasion this goes amusingly wrong independent.co.uk/news/uk/politics/…
    – origimbo
    May 22, 2018 at 20:59

1 Answer 1


It's notoriously hard to prove a universal negative, but I believe it's very possible there exists no such study for UK local elections. Among other issues:

  1. UK local elections consist of thousands of individual contests, some of which spread over multiple levels of government and multiple different voting systems.

  2. Small voting pools are complicated to poll for, requiring larger samples in comparison to the size of the population to achieve statistically significant results. At the same time, there are added logistical details in sampling the eligible voters, and organising a presumably doorstep polling campaign. Polling would be necessary, since correlations alone don't prove causation.

  3. What aggregate polling there is consistently shows a strong tribal effect in voting patterns, as well as the influence of the national picture and pre-existing local history. Since you imply your friend wasn't working from a template supplied to him or her, I will assume he or she was standing as an independent, or for a minor party, who have been especially impacted by the swing back towards two party politics at the national level.

You must log in to answer this question.