Who? Why? Johnson? And When?

A look at Boris Johnson’s precarious political position, and cautionary tales about ‘hypothetical’ polling

Two for the price of one in today’s newsletter, as we look back on another week of chaos for Boris Johnson’s government and ‘ahead’ to hypothetical new Tory leaders

Johnson battles on through storms both near and far

As we look ahead to another week of potential political turmoil for Boris Johnson and his beleaguered government, it is worth taking stock on where his position is right now with both the public and his own party.

Firstly, Johnson’s attempts to calm the mood and arrest the situation in the wake of the Sue Gray report with promises of reform and change have not struck any kind of chord with the public.

According to YouGov snap polling released shortly after the prime minister’s address to the House of Commons on Monday afternoon, still around 2/3 of the British public (63%) believe that Johnson should resign. That represents a change of -1 points on January 17th (in statistical terms, absolutely no change at all).

The public are, for now, completely unconvinced that Johnson’s proposed changes – many of which we began to see this week with the resignations of key staff including Martin “BYOB” Reynolds and communications chief Jack Doyle, followed by the appointments of Steven Barclay and Guto Harri – will amount to anything all that substantial in terms of change.

Just 5% think have ‘a great deal’ of confidence that the prime minister’s plan will make substantial changes, while 7 in 10 have little to no confidence in that at all.

That said, while the public generally do not believe the prime minister’s accounts and recollections of lockdown-busting Downing Street parties, they also do not believe that he is solely to blame for what went on.

Though around one-in-four (38%) believe that Johnson is most responsible for the workplace culture at Number 10, the same proportion (39%) think that both he and his staff are equally responsible. Only 13% believe his staff are more responsible.

Those figures suggest that a majority of the public could be convinced, at least to some degree, that action has been taken and improvements will be made through resignations and personnel changes.

Whether this will be enough to placate the mood among the British public for Johnson to go however is a different question entirely.

The Big Dog barks on, for now

As far as Johnson’s position within his own party goes, he has reason to be both cheerful and gloomy about the days and weeks ahead. Firstly, despite quite some speculation and expectation at the beginning of the year when the ‘party gate’ storm clouds rolled in, Boris Johnson has not as of yet been subject to a vote of no confidence by his MPs.

Operation “Save Big Dog” achieved success in stemming the flow of letters to 1922 Committee Chair Sir Graham Brady – and according to a few sources, succeeded in persuading a handful of the 2019 intake to rescind theirs – and pushing out new communication lines focusing on the government’s record of delivery on Brexit and vaccines, and their desire to “move on” and “get on with the job”.

Around a week ago, my ‘Post-it Note’ figure for letters sat on Brady’s desk was 25. Most sources and commentators believe that number to be somewhere between 35 and 45 now. The figure could even be closer to 50, according to some, meaning the prime minister is potentially only one more gaff or negative story away from facing down his own MPs in a confidence ballot.

It remains unlikely though that many of those plotting to oust Johnson would want, ideally, to trigger the vote before this year’s local elections and/or the conclusion of the Met Police investigation (and subsequent publication of the full Sue Gray report) into the Downing Street parties.

There is a sense that, as it stands, with the police investigation masking vital party gate details and the coming political fallout of rapidly rising inflation (which none of his potential successors would enjoy walking straight into) acting as a double-shield, the prime minister could well survive a no-confidence vote if it were held in the coming days.

In the meantime, put down the hypothetical polling

As talk of a Conservative leadership contest rumbles on, we’ve seen an increasing amount of polling released comparing the fortunes of various prospective Tory leaders against Keir Starmer in a variety of electoral contests (e.g., the ‘Red Wall’, an election tomorrow, in 2024, and so on).

Most recently, the fantastic long read by Tim Shipman, Caroline Wheeler, and Gabriel Pogrund, reported on a Focal Data poll which claimed that Sunak had the ‘best chance’ of reaching voters in the ‘Red Wall’, compared to Johnson, against Keir Starmer at the next election.

The piece also references a poll supposedly carried out by Survation for the Labour Party – which Chief Exec Damian Lyons Lowe denies actually happened (my theory is that a bit of wire crossing has gone on, and that the poll was done but fielded by someone else) – which apparently found that Starmer would lose an election against any prospective Tory leader other than Boris Johnson.

Elswhere, the i reported last year on a poll carried out by Opinium, which showed Labour would pull 18 points ahead if it were Michael Gove leading the Conservative Party. Potential electoral disaster for 2024.

But while these polls do tell us something about the current political scene, it’s not actually what’s written on the tin.

In truth, prospective, hypothetical polling such as this tends to be highly inaccurate in terms of predicting actual results or voting intention, and are actually largely just a test of name recognition with a bit of personal favourability fixed in.

So, not totally useless, but not very useful either.

On the first, some avid readers may recall a poll from ICM in August 2018, suggesting that replacing Theresa May with Boris Johnson would likely lose the Conservatives the next election.

What happened of course is that the Conservatives did replaced May with Johnson, and then went on to win their biggest majority since the Thatcher years.

Across the pond, others might recall the extensive polling in 2015, during the heat of the primary contests, suggesting that Hilary Clinton would come out handily on top if she were pitched against Donald Trump in the forthcoming 2016 Presidential election.

One particular poll published by NBC/WSJ had Clinton 10 points ahead of the candidate who would eventually be her election rival, with each of Carly Fiorina, Ben Caron, and Jeb Bush either at level pegging with or ahead of Clinton in prospective vote intention. The recommendation from the polling to the Republican Party was clear – pick anyone other than Trump.

What happened of course is that the Republicans did pick Donald Trump in 2016, and he went on to win the election. Rather than a 10 point lead, Clinton ended up just 2 points ahead of Trump in the popular vote, and fell to a substantial Electoral College defeat.

A much better approach to gauging public mood toward prospective prime ministers at this stage in an election cycle – namely, hypothetical and extremely uncertain – is to simply poll name recognition and favourability.

Between these two, we get a strong sense of each candidate’s respective positions ahead of leadership and (later) nationwide elections. Respectively, they tell us (a) how many people have actually heard of the candidate, and (b), of those who have heard of them, what is the general consensus about them.

Hypothetical leadership polling at this stage of an election is essentially a mirror of name recognition anyway. We so often see in the US for instance that defeated primary candidates from the last cycle more often than not lead (or are in the top two or three) in early polls for the following primaries. That’s simply because people have heard of them.

Equally, here in the UK, we often find prospective leaders of whom the public has not heard very much will often do far better than more established politicians with whom they are more familiar with and have formed an opinion on.

For instance, Penny Mordaunt will more than likely poll much more positively as a prospective Conservative leader than Michael Gove. The reason being is that far fewer people have heard of Mordaunt, and so far fewer people have negative opinions of her than they do of Gove.

We can see this clearly in YouGov’s ‘popularity and fame’ data. According to that, Mordaunt is known by around 1/3 of the British public, liked by 13% of them, and disliked by 10%. Gove on the other hand is known by almost 90% of the British public, is liked by 18% of them, and disliked by 52%.

If a candidate has high public recognition, then favourability is a probably a half decent indicator of how the public might receive them as a leader. If it is low, then their favourability is not a useful indicator at that stage but the evolution of their favourability over time – and as they become more recognised – then becomes a good indicator.

But really, only once a candidate is installed as leader does the question of how they might perform at a General Election become relevant. Even then, it only becomes useful once an election draws near – polling about Keir Starmer from two years ago, for instance, would have been a terrible predictor of his current leads against Johnson in ‘best Prime Minister’ and personal favourability.

And it is also only relevant in comparison with (or relative to) the ratings of their political opponent. For instance, Boris Johnson’s personal ratings were not at all impressive in the lead up to the 2019 General Election, floating at around -20, but they were streets ahead of the awfully unpopular Jeremy Corbyn, who sat at about -50.

In short, the vast majority of the public do not simply think in such abstract terms as to consider how they would vote under different combinations of leaders. They aren’t able to answer questions such as “If candidate X was leader, how would this affect you in way Z” because they do not spend their time forming views on questions like this in their day to day life. They have better things to do!

As such, any questions about prospective leaders are essentially a combination of name recognition and favourability, especially this far out from a scheduled election.