Reuters logo
What is the YouGov model which estimates May could lose majority?
May 31, 2017 / 10:54 AM / in 5 months

What is the YouGov model which estimates May could lose majority?

Britain's Prime Minister Theresa May speaks during a Q&A at Cross Manufacturing Company in Odd Down in Bath, Britain, May 31, 2017. REUTERS/Leon Neal/Pool

LONDON (Reuters) - Prime Minister Theresa May could lose control of parliament in Britain’s June 8 election, according to a projection by polling company YouGov, raising the prospect of political deadlock just as formal Brexit talks begin.

What is the YouGov model, how does it work and what does is predict?

WHAT DOES THE YOUGOV MODEL PREDICT?

Party

YouGov estimates

2015 election

Conservative Party

310

331

Labour Party

257

232

SNP

50

56

Northern Irish parties

18

18

Liberal Democrats

10

8

Plaid Cymru

3

3

Green

1

1

Other

1

0

UKIP

0

1

HOW DOES IT WORK?

Every day YouGov interviews approximately 7,000 panellists about their voting intentions for the June 8 election. About 75 voters per constituency per week are asked their opinions. Over the course of a week, data are collected from around 50,000 panellists.

That data is fed into YouGov’s model, which factors in interview date, constituency, voter demographics, past voting behaviour, and other respondent profile variables. The model estimates the probability that a voter with specified characteristics will vote Conservative, Labour, or some other party.

Then the model uses data from the UK Office of National Statistics, the British Election Study, and past election results, to estimate the number of each type of voter in each constituency.

“Combining the model probabilities and estimated census counts allows YouGov to produce a fairly accurate estimate of the number of voters in each constituency intending to vote for a party on each day,” YouGov said.

WHAT IS IT CALLED?

YouGov is using a new technique called Multilevel Regression and Post-stratification (‘MRP’ for short) to produce estimates for small geographies (local authorities for the EU referendum, states in the 2016 American Presidential election, and parliamentary constituencies for the 2017 General Election).

WHAT ARE THE LIMITS OF THIS MODEL?

- the samples in each constituency are too small to be reliable by themselves and are subject to more than just sampling error. To compensate for small sample sizes, YouGov relies on a model that pools data across constituencies. This uses data from panellists who live in other constituencies to augment the small number of actual interviews conducted in a constituency.

- The model is based on the fact that people with similar characteristics tend to vote similarly, but not identically, regardless of where they reside, YouGov said.

- Using MRP, YouGov have classified constituencies as safe, likely, or leaning to a party or as a toss-up.

HAS IT BEEN USED BEFORE?

Yes. YouGov said its MRP model in the 2016 EU referendum consistently showed that more voters favoured leave than remain, and that Hillary Clinton would win the popular vote in the 2016 US Presidential election by a narrow margin, but that midwestern battleground states were too close to call.

WHO MADE THE MODEL?

The model was developed by Professor Ben Lauderdale of the London School of Economics in conjunction with YouGov’s Data Science team, headed by Doug Rivers of Stanford University, YouGov said.

The data are streamed directly from YouGov’s survey system to its database. From there, the models are fit using Hamiltonian Monte Carlo with the open source software Stan.

Stan was developed at Columbia University by Andrew Gelman and his colleagues, with support from YouGov and other organisations.

YouGov will be updating the model estimates on a daily basis.

Reporting by Paul Sandle and Robin Gillham; Editing by Guy Faulconbridge and Raissa Kasolowsky

0 : 0
  • narrow-browser-and-phone
  • medium-browser-and-portrait-tablet
  • landscape-tablet
  • medium-wide-browser
  • wide-browser-and-larger
  • medium-browser-and-landscape-tablet
  • medium-wide-browser-and-larger
  • above-phone
  • portrait-tablet-and-above
  • above-portrait-tablet
  • landscape-tablet-and-above
  • landscape-tablet-and-medium-wide-browser
  • portrait-tablet-and-below
  • landscape-tablet-and-below