Phantom Polling

Anthony B. Masters
4 min readMar 9, 2019

--

On Bloomberg, Iain Duncan Smith MP (Conservative, Chingford and Woodford Green) cited supposed polling which showed that “the vast majority want to get out [of the European Union] now, even if they voted Remain”.

This article compares the former Conservative leader’s claims about polling to actual polling results. Mr Duncan Smith has severely misrepresented public opinion.

In short

“Vast majority want to get out now”: Opinium polls suggest the public are split on their next option, with less than 3 in 10 supporting leaving with no deal. No Deal typically comes second in three-way referendum questions.

“Happy”: YouGov have been the only company to ask about personal happiness — and found that No Deal made more people unhappy than happy.

“Same in my constituency”: YouGov estimated public opinion in each constituency, finding Remain ranked first in Chingford and Woodford Green.

“The vast majority”

On Bloomberg, Iain Duncan Smith MP was interviewed, and made numerous claims about what polling of British public opinion showed:

If you look at the polling, it’s quite clear that the vast majority want to get out now, even if they voted Remain — just get on with it. All the polling tells us categorically. It’s the same in my constituency: the majority are happy to just go out without a withdrawal agreement. That gets the highest scoring of the lot.

The main two claims made by Mr Duncan Smith on Bloomberg were:

  1. The vast majority want to leave the EU without a withdrawal agreement;
  2. Most people would be happy in that scenario.

Actual polling — rather than phantom polling — suggests that these claims about public opinion are untrue.

Opinium, on behalf of The Observer, have frequently asked people what they want to happen next, after the proposed agreement had been rejected by Parliament.

The public are split. Of the six offered options (including don’t know), leaving the EU without a deal has the highest share. In each of these nine readings, less than 3 in 10 people supporting leaving without a deal, which is far from a “vast majority”.

A Remain-Deal “public vote” is the consistent second option. (Source: What UK Thinks EU)

Some polling companies have asked people how they would vote in a hypothetical ballot with the options: leave with the proposed agreement (‘Deal’), leave with no deal (‘No Deal’), or remain in the EU (‘Remain’).

In the 11 polls by British Polling Council members, Remain has a consistently large lead over No Deal.

The Deal option is generally placed third, in these questions about top preferences. (Source: What UK Thinks EU)

Happy with no deal?

In four polls since August 2018, YouGov has asked people how happy they would feel if the UK left the EU without any sort of trade deal:

If Britain leaves the EU and does not sign any new trade deals with the EU (meaning no automatic access to European markets and tariffs or other barriers on many British exports) but has full control over its borders, does not have to follow any EU regulations and does not have to contribute any money to the EU, would you personally be happy or unhappy with this outcome?

Only a minority say they would be fairly or very happy with this outcome, though the difference has narrowed in the latest reading.

A relative majority would be unhappy with leaving the EU in the absence of a trade deal. (Source: What UK Thinks EU)

What is popular in Chingford and Woodford Green?

The former Secretary of State also made a claim about what was popular in his constituency. It is normally not possible to evaluate such claims, but YouGov have estimated popularity of the EU exit options in each constituency.

Between 27th November and 4th December 2018, YouGov asked 20,910 Great British adults via its internet panel to rank these three options: Deal, No Deal, or Remain.

Using this data, Ben Lauderdale and his colleagues build a model of individual’s likelihood to prefer each ranking, based on their demographic characteristic (such as how young men would vote). These individual likelihoods are then translated onto constituencies. (The statistical method is called Multi-Level Regression with Post-Stratification, or MRP.)

Remain is the first preference in an estimated 600 constituencies. (Source: YouGov)

We can look at the estimated opinion in Chingford and Woodford Green. Like most other constituencies, Remain wins on first preferences (with an estimated 47%). If we were to knock out the most unpopular of the three options (No Deal, in the case of that constituency), and redistribute their second preferences, then Remain wins again.

Remaining in the EU is estimated to be popular in Chingford and Woodford Green. (Source: YouGov)

Lastly, if we were to compare Remain versus Deal, Remain versus No Deal, and Deal versus No Deal, then Deal is estimated to win two of these contests. In political science, Deal is called the Condorcet winner in Chingford and Woodford Green.

Like other data sources, claims that “polling tells us categorically” should be made in reference to actual polls, not phantom polling.

--

--

Anthony B. Masters
Anthony B. Masters

Written by Anthony B. Masters

This blog looks at the use of statistics in Britain and beyond. It is written by RSS Statistical Ambassador and Chartered Statistician @anthonybmasters.

No responses yet