Diagnoses, Opinion and Flags

What happened to public opinion after the Prime Minister’s test result?

In a popular post on Twitter, The Observer writer Carole Cadwalladr claimed:

the net effect of our prime minister catching COVID-19 was that it prompted a surge of patriotic support.

Opinion polling does not substantiate this belief. The balance of favourable views appeared to fall after the Prime Minister’s test result. In some polls, the PM’s hospital stay may have induced a temporary rise in his personal approval.

A “surge” of support?

Following the US President’s positive test result for SARS-CoV-2, Carole Cadwalladr asserted:

A reminder to all Americans that the net effect of our prime minister catching COVID-19 was that it prompted a surge of patriotic support. From which he emerged with renewed popularity. Which enabled him to tear up key functions of the state

Over 44,000 users shared this claim. This article will look at government and personal approval measures from some companies.

In the UK, numerous major events happened in quick succession:

  • 11th March: The World Health Organisation declares a global pandemic.
  • 16th March: Social distancing guidelines begin. There are recommendations to avoid non-essential travel and work from home.
  • 18th March: Schools close to all pupils, except children of key workers.
  • 23rd March: In a televised address, the PM issues a stay-at-home order.
  • 27th March: The PM and Health Secretary both receive positive test results for the virus.
  • 5th April: The Prime Minister enters to hospital with COVID-19.
  • 12th April: The Prime Minister leaves hospital, after spending time in intensive care.

Looking at monthly polling may mean entangling any of these effects. There are at least four polling series with weekly or daily estimates. I include: YouGov, Savanta ComRes, Opinium, and Morning Consult (US). Surveys provides estimates, subject to many sources of potential error. There are plausible ranges around each figure: it could be somewhat higher or lower.

These polling companies ask questions with different wordings, and use different weights. Those choices mean there are systematic differences between companies — house effects.

The four polling series tell similar stories:

  1. People had more favourable views of the Prime Minister after the government response. This is before the PM receives his diagnosis.
  2. Net approval then remains static or falls a small amount after the COVID-19 diagnosis. Approval appears to rise after hospitalisation.
  3. That rise was temporary — dissipating by the start of May. In that month, favourable opinion of the Prime Minister continued to decline.

Four series

On behalf of The Times and Sunday Times, YouGov ask their internet panel:

Do you have a favourable or unfavourable opinion of Boris Johnson?

In their weekly tracker, YouGov estimate Johnson’s net favourability to be -3 on 17–18th March. Net favourability means the share of favourable answers minus the unfavourable share.

That point estimate surges to +20 on 22–23rd March. There is a fall to +17 on 1–2nd April. Favourability increases to +29 on 10–14th April — whilst the PM is in hospital. On 5–6th May, YouGov’s net favourable estimate was +16.

Image for post
Image for post
The dates after 27th March should all say April. (Image: Twitter/YouGov)

Savanta ComRes ask their internet panel respondents:

From what you have seen or heard to what extent do you approve or disapprove of how each of the following public figures are handling the outbreak of COVID-19?

A similar shape of approval over time forms:

Image for post
Image for post
These are weekly polls. (Image: Savanta ComRes)

Morning Consult run an internet panel. Again, their net approval follows the same pattern over time:

Image for post
Image for post
The exact wording is unclear, but it is a question about approval. (Image: Morning Consult)

Opinium ask their approval question on behalf of The Observer:

To what extent do you approve or disapprove of the way Boris Johnson is handling his job as Prime Minister?

Due to the timing of the survey waves, the pattern is less clear. Their internet panel polling suggests his net approval was +29 on 26–27th March. This fieldwork overlaps with the PM receiving a positive test.

The next reading showed no change: +29 on 1–3rd April. That figure then falls by a single point: +28 on 15–17th April. That small change could be sampling error.

Rallying to the flag

YouGov’s series suggests rising support for the government preceded Johnson’s test:

Image for post
Image for post
The hospitalisation did not appear to have an effect either. (Image: Twitter/YouGov)

As Matt Singh (Number Cruncher Politics) writes:

there was a surge, but before, not after, Johnson went into quarantine (and subsequently into hospital).

Image for post
Image for post
Vote intentions do not show major changes after the positive test. (Image: Number Cruncher Politics)

In certain crises, events drive popular support for incumbent governments. American political scientist John E. Muller called this the ‘rally to the flag’ effect. As Prof Will Jennings (Southampton) describes:

In all countries leaders have been a central figure in the government response. The existential threat posed by the virus, combined with radical restrictions on movement that have been implemented, growing pressure on health services and severe shocks to the economy, all are features consistent with the dramatic and sharply-focused events noted by Mueller.

Morning Consult found many national leaders enjoyed a surge in popular approval:

Image for post
Image for post
This is net approval, so approval minus disapproval. (Image: Morning Consult)

Another issue is: can we generalise the change in Johnson’s approval (or lack thereof) to the United States?

Social media can join people in different countries through language. The politics of those countries are not identical.

The United States has a presidential system of government. The United Kingdom has a parliamentary democracy. The two countries are at different stages of their electoral cycle. The US counts votes from 3rd November. The UK had a general election in December 2019. The two events are at different phases of the pandemic. The US incumbent has a polling deficit, whilst the Prime Minister had a polling advantage.

Many major events occurred in quick succession. The Prime Minister’s improved approval preceded his positive test result. This follows people rallying to the flag in times in crisis. There appears to be a modest boost following hospitalisation — but it was temporary.

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

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store