The last 3 general elections have seen some significant polling errors. In 2010, the Lib Dems were significantly overestimated, in 2015 the Conservatives were underestimated and last year saw the largest ever underestimate in the Labour vote. Whilst these errors suggest that the polling industry is struggling with general elections these days, a natural question to ask is “are all pollsters equally bad or are some better than others?”
Ahead of the 2017 general election, I predicted that the opinion polls would be wrong again and that the Conservatives lead over Labour would be underestimated by 2.6%. I based this on data provided by Mark Pack who has systematically recorded every opinion poll published since 1945. In the event, I was right that the polls would be wrong but instead of an error favouring the Conservatives, the polls recorded the largest ever underestimate of the Labour vote. As a result, election forecasters were blindsided yet again and the result was a hung parliament which few saw coming.
This is my last update of the opinion before election day on June 8th. I will use my analysis of these polls to update my 2017 General Election Seat Predictions and you should read that post in conjunction with this one.
The latest situation is that the Conservatives now only hold a 7% lead over Labour which is down 2% from last week and only just above what they had in 2015. Labour’s vote share has recovered significantly to narrow the Conservatives lead and Labour are now capable under some scenarios of depriving the Conservatives of a majority. However, it has now become clear that pollsters are dividing into two groups and I have written a separate post that explores the implications of this and I strongly recommend you read that. This post reports on all posters combined rather than separate blocs.
Unlike the 2015 general election when the polls were essentially static (& wrong) throughout the election, the 2017 general election has seen some of the most extraordinary volatility in the polls that I can remember. If you are a Conservative supporter, the narrowing lead over Labour must be leading to anxiety and changed underwear. If you are a Labour supporter, you are probably starting to dream “can we? will we?!” It doesn’t help that your state of mind will depend on which poll you are reading and your memories of the pollsters’ failure in 2015 so how can you make sense of what is going on. I will show you how in 5 steps and to heighten the drama, I will leave the punchline to the end!
As I have explained, I track the regional breakdowns to produce this chart. Every now and again, a full on poll takes place to explore a region in more detail and this allows me to check how good chart R1 is.
My wife is American and so it should be easy to guess what we were talking about on the morning of 9th November 2016. Donald Trump’s victory in the US Presidential election was a surprise to many people and prompted much discussion on the similarities between Trump voters and the Leave voters in June. However, my wife remarked that people may be looking at this the wrong way round and perhaps the correct question to ask is whether there is greater similarity between Clinton & Remain voters.
Identifying similarities and differences between groups of people is a cornerstone of the field of market research known as customer segmentation. It is one of my favourite areas of statistics and can be used regardless of whether the data comes from a survey or from customer records. When my wife posed her question I immediately thought of 2 ways I could answer this using segmentation methods.
- Look at how people feel (their sentiments) which is what this post is about.
- Look at how people voted (their behaviour) which I will cover in another post “Who has more in common? Leave & Trump voters or Remain & Clinton voters? Analysis of voting behaviour”
Since the UK voted to leave the EU on 23rd June 2016, there have been 3 contested parliamentary by-elections (Witney, Richmond Park, Sleaford & North Hykeham) and one uncontested by-election (Batley & Spen which was the late Jo Cox’s seat). Many commentators have analysed these results to see how the referendum result has impacted on parliamentary voting intentions. Whatever voter dynamics are revealed, it is reasonable to assume that they are likely to influence future by-elections. In late October 2016 just after the Witney result, I realised it could be possible to build a by-election model by combining two sources of data.
- My own estimates of the Leave & Remain votes in each of the 650 parliamentary constituencies where I calculated that 400 out of 650 seats voted Leave.
- My interpretation of the Lord Ashcroft “exit poll” carried out on 21st to 23rd June 2016 and published immediately after the results were announced.
At the time, I described my by-election modelling approach in a youtube clip and that is worth listening to. I have made some changes to my model since then so this post is the most up to date version of my model. I will illustrate the basic principle using the Witney by-election (David Cameron’s former seat) of 20th October 2016 where the top line numbers are: