Best and Most Controversial Pollster, YouGov's Latest Results

-edited

YouGov's latest poll shows another pickup for Labour with support for Tories unchanged.

YouGov's poll on Nov 25-26 had an 11 percentage point advantage for the Tories.

  • Tories 43
  • Labour 32
  • Lib Dem 13

The +2 pickup is thus undecided voters breaking for Labour.

9 points (if accurate) is still easily enough for a Tory majority, assuming it holds. I believe it will.

Five Point Synopsis

  1. Tory support stabilized at 43%. All the recent polls have the Tories between 41% and 43%. The hung parliament in 2017 had a Tory advantage over Labour of 43-41. There is more tactical voting this time but 9 points is more than enough to cover it. Even 6 points could easily be sufficient.
  2. Brexit party support has collapsed. Perhaps the remaining support costs the Tories a few seats because what support remains is concentrated.
  3. The Green party was never much of a factor to begin with. But there is next to nothing left for the Liberal Democrats or Labour to pick up from the Greens.
  4. The Liberal Democrats stabilized at 13%. That is probably close to core support. If so there is nothing left for Labour to pick up from Liberal Democrats. If support for Liberal Democrats picks up, it will be at the expense of Labour.
  5. Labour needs a huge pickup from the undecideds, and they have to actually turn out to vote.

ComRes Has Tories Gaining Ground

ComRes has the Tories gaining from its previous poll.

Opinium Still Has Monster 15-point Tory Lead (Down from 19)

BMG

If you care to believe BMG, the Tory lead is down to 6.

Why the Discrepancies?

BMG has the spread at 6, YouGov at 9, and Opinium at 15.

Matt Singh asks and answers Why are the Polls Telling Such Different Stories?

How so these consistent differences arise? Contrary to the widely-peddled myth, it has nothing to do with the editorial leanings of the newspaper publishing it.

It’s easiest to start by explaining what polls do. Polls estimate public opinion across a population by taking a sample of the population. The statistical basis for this is that the sample is either selected randomly – and therefore by definition representative of the population – or if not random, that can be made representative of the population.

Since there are many ways in which methodologies can differ, and many of them interact with one another, it can be difficult to unpick exactly why two or more pollsters consistently arrive at different estimates of public opinion. But sometimes the source of the variation can be inferred.

In 2017, for example, final polls varied wildly on the size of the Conservative lead. In this case, the primary reason was clear. Those showing narrow Tory leads over Labour were basing their turnout assumptions on how likely people said they were to vote. Those showing wider leads were using complex turnout modelling instead. Some of these models went very over-the-top, leaving too few Labour voters among those expected to vote.

In 2019, there again seems to be a clear explanation for at least part of the discrepancy. Among the possible weighing variables mentioned above, many UK pollsters weight their samples to ensure that the way their respondents voted in the last election matches the result. In theory, this makes complete sense – a representative sample should match the result.

In practice, it’s not so simple. Though it may sound strange to political anoraks, a lot of people are very, very bad at remembering which way they voted, even in the recent past. A number of pollsters have conducted experiments where they interviewed people who had said just after the 2017 election that they had voted Labour, and asked them once again how they had voted in 2017. In one case, about one in five, or about 8 per cent of all 2017 voters, thought that they had voted for a party other than Labour.

So it follows if we have the right proportion of actual 2017 Labour voters (41% of the Great Britain vote) we would have a much lower proportion of those remembering having voted Labour last time. If we were to weight people’s recalled 2017 vote to the result, we would then end up with too many 2017 Labour voters.

There are a few ways to deal with this. Some online pollsters, including YouGov and Opinium, past vote weight using 2017 votes collected at the time, thus taking memory out of the equation. Others, such as Kantar and Deltapoll, weight using recalled past vote, but to a target adjusted for false recall. And still others, namely Number Cruncher and Ipsos MORI, don’t rely on past vote weighting at all.

YouGov's Model

Please consider How YouGov's 2019 General Election Model Works.

In recent UK and US elections, YouGov has been using a technique called Multilevel Regression and Post-stratification (or 'MRP' for short) to produce estimates for small geographies.  These include local authorities for the EU referendum, states and congressional districts in US elections, and parliamentary constituencies for UK elections.

The idea behind MRP is that we use the poll data from the preceding seven days to estimate a model that relates interview date, constituency, voter demographics, past voting behaviour, and other respondent profile variables to their current voting intentions. This model is then used to estimate the probability that a voter with specified characteristics will vote Conservative, Labour, or some other party. Using data from the UK Office of National Statistics, the British Election Study, and past election results, YouGov has estimated the number of each type of voter in each constituency. Combining the model probabilities and estimated census counts allows YouGov to produce estimates of the number of voters in each constituency intending to vote for a party.  In 2017, when we applied this strategy to the UK general election, we correctly predicted 93% of individual seats as well as the overall hung parliament result.

Best but Most Controversial Pollster

Wired explains How YouGov Became the UK's Best but Most Controversial Pollster.

On the evening of June 8, 2017 Stephan Shakespeare, the 62-year-old co-founder and CEO of British polling company YouGov, was nursing a lager in The Hairy Canary, a Brussels pub not far from the European Commission’s HQ. It was a tense moment for Shakespeare: the results of the 2017 UK general election were about to be declared, and YouGov had predicted that prime minister Theresa May would lose her majority and end up with a hung parliament. It was the only pollster with such gloomy an outlook for the Conservatives, at a time when the general consensus was that May would romp home to victory. Shakespeare normally drinks wine, but picked a long drink because he was expecting a long night.

“It was very hard to be out there with a number that was so different from everybody else’s,” Shakespeare said back in 2017. “We were very exposed: we were calling every single seat using a new methodology.”

On the evening of June 8, 2017 Stephan Shakespeare, the 62-year-old co-founder and CEO of British polling company YouGov, was nursing a lager in The Hairy Canary, a Brussels pub not far from the European Commission’s HQ. It was a tense moment for Shakespeare: the results of the 2017 UK general election were about to be declared, and YouGov had predicted that prime minister Theresa May would lose her majority and end up with a hung parliament. It was the only pollster with such gloomy an outlook for the Conservatives, at a time when the general consensus was that May would romp home to victory. Shakespeare normally drinks wine, but picked a long drink because he was expecting a long night.

“It was very hard to be out there with a number that was so different from everybody else’s,” Shakespeare said back in 2017. “We were very exposed: we were calling every single seat using a new methodology.”

YouGov turns to online panels comprising a million people in the UK – and millions more internationally – whose members are constantly consulted for their opinions on everything from the news of the day to which brand of bread they buy. (When taking a poll on 26 November, I was asked everything from my opinion on the Chief Rabbi’s fulmination against Labour to what feelings I had about Chelsea FC, and whether I eat to live or live to eat.)

People who participate in YouGov panels are “paid” in points, usually receiving around 50 points for a daily poll. Once they reach 5,000 points, they can cash it out for £50.

Up until YouGov arrived on the scene, market researchers were dour-looking men and women, clipboard in hand, who would pelt busy shoppers with questions. The participants would be chosen at random, and would be complete strangers, proffering their opinions to the survey organisers for the first and possibly last time. Sometimes, as a sop to technology, researchers would call up people in their homes at random and ask them the same questions.

YouGov proposed an alternative method: remaining in contact with survey-takers through the internet, constantly questioning them about their preferences, in order to better track the changing tastes of consumers over time.

That trove of personal information – and the ability to link someone’s tastes in household items to their likelihood to vote a certain way – is meant to sort out the one big problem pollsters have: sampling.

Representative samples are difficult to achieve, particularly across all 650 constituencies in the UK. “The thing people get most het up about is sampling,” explains Eric Harrison, a senior research fellow at City University London’s sociology department. “Getting a representative sample has been increasingly difficult due to people’s reluctance to answer questions on telephone polls, and one can’t rely on voluntary panels to be representative,” says David Spiegelhalter, a statistician at the University of Cambridge.

Using MRP accepts that challenge, and deploys some statistical fixes to process the data after collection.

YouGov’s MRP is meant to counteract the fact that the people being polled are self-selected, by removing the likelihood that respondents are more politically-literate or skewed to one side of the political spectrum . “We have our MRP running now, and it keeps getting adjusted,” says Shakespeare. One judgement update they’re having to make this election is around the likelihood of tactical voting. “We collect something like 10,000 interviews every single day just for this.”

Most Likely Outcome

People can and will believe what they want. Here, the most likely outcome, for several reasons is something on the order of an 8-10 point Tory advantage.

  1. 8-10 points is near the average
  2. 9 points is what the pollster who had a 93% seat projection as well as a correct hung parliament call in 2017 predicted in advance.
  3. All the polls suggest there is little room left for further tactical voting. It's already reflected in the polls.

At this point, it will take a huge swing in undecided voters making up their mind not only to vote, but to vote for Labour.

It's possible, just not the most likely outcome.

Mike "Mish" Shedlock

Comments (17)
No. 1-6
Deep Purple
Deep Purple

Thanks for the summary. The polls are a mess but they show a tighter competition than a few weeks ago. We will see if it goes further or not.

By the way, I don't think that tactical voting can be fully considered in advance. It is a quite complicated thing.

TheLege
TheLege

What does avidremainer think?

Anda
Anda

What I reckon is that when voters see there is a good lead for their party, they are less likely to vote, and because this is known those who are definitely going to vote are more likely to. Therefore those who don't vote are going to increase because they find that a lot of definite voters is off-putting as it shows unconsidered bias in human behaviour, so they definitely won't vote as well. Those who were undecided on whether to vote will become even more so as the nation moves towards a vote/no vote paradigm, and eventually will feel completely trapped because whatever they do they will be placed in one category or another. This will bring out a sympathetic response from definite voters and many will also declare themselves as undecided out of compassion for the trapped. As voter turnout estimates plummet due to this, the political discourse will turn to questions of social disintegration as a reflection of government ineptitude and draconian measures to encourage voting will be agreed by a cross party majority. A parliamentary bidding war will errupt when deciding the limit of payments able to be offered to potential voters for casting a vote. Questions of inequality in terms of distance to the booth and compensation for that will run as quiet sidelines to the main arguments surrounding fiscal leverage and if there is any point not paying maximum to stimulate the economy, and even should everyone get paid so as not to discriminate against non-voters and those with intermittent media access that might find their decisions conflicted by lack of information. The sum result of this will be utter confusion as non-voters will vote because otherwise they would be being paid for not voting, and dedicated voters will not vote because they feel their vote would be seen as for pay. As government tries to counteract that trend people will get so upset at having their payment withdrawn that they will form large crowds in front of their favourite shopping centres , which will be found empty of stock after the crowds disperse.

lamlawindy
lamlawindy

Has YouGuv polled any specific constituencies? I'd love to see a poll on Doncaster North (Ed Miliband) & Pontefract/Castleford (Yvette Cooper). Both areas were around 70% "Leave" in 2016; I gotta think that a chunk of their constituents would be voting against Labour.

footwedge
footwedge

I don't have a dog in this hunt but I have a couple Brit retired military friends who apparently do and they have a some what less favorable view of the honorable PM. My biggest problem is his freaking hair - geez, don't they have barbers in the UK! (Not unlike another world leader with similar proclivities I guess.)

Mish
Mish

Editor

YouGov has polled every region. Unfortunately, the data they post is very stale.

I will crunch some numbers if I can find a new poll that has all the regions.

YouGov only broke out London in the latest poll. None of the others broke out anything that I can find