For some time now, it has been apparent that something is happening in the world of politics. I don’t mean the stupid office parties during lockdown. What I mean is that large numbers of people are hugely disappointed in Boris and his government – and are beginning to feel completely disenfranchised. The dismal state of Her Majesty’s Loyal Opposition has only added to the feeling that many are expressing: that of electoral impotence. The total chaos and constant changes of mind; of promises which are made and then broken; of disruption to the economy and education; an increasingly incompetent NHS; and an overwhelming feeling that we are all being lied to on a daily basis, have contributed to this discontent.
Whilst there have been lots of well-publicised opinion polls suggesting that everything in the garden is lovely, these polls cannot be trusted. The idea that a sample size of 1000 carefully chosen online panellists are going to represent the views of 65 million people is, of course, risible.
But we now have a small collection of three recent by-elections wherein actual electors have expressed their views at the ballot box. Taken together, they amount to nearly 98,000 votes cast. These are the only verdicts upon government performance which are worth paying any attention to.
The three by-elections concerned are (with the dates of the election): Chesham and Amersham (17th June 2021); Old Bexley and Sidcup (2nd December 2021); and North Shropshire (16th December 2021). Until their respective by-elections, all of these constituencies have been regarded as Conservative strongholds. At the 2019 General Election they had Conservative majorities of between 29% (Chesham) and 41% (both Bexley and N. Shropshire). All have been held by the Conservatives for decades.
These characteristics indicate that these three constituencies are very similar. The relative stability of these constituencies in electoral terms, and then the sudden changes recently, suggest that all three are worthy of study together. This article looks at these by-elections with particular reference to the voting behaviour of people who would normally vote Conservative but have chosen to withhold their votes from Boris Johnson’s government. It also looks at the sources of voters who appear to have switched in order to maximise their protest.
The following article is nearly 4000 words long and contains nine tables. Each table is accompanied with an explanation of the logic of the calculations. However, those who find this tedious reading can jump straight to the Summary section which is to be found towards the end of the article, followed by the Discussion and Postscript. The basic numbers are all taken from the Wikipedia articles about each by-election here, here and here.
One place to start analysis is to look at turnout percentages:
|Registered Electors (2021)||% Turnout 2019||% Turnout 2021|
|Chesham & Amersham||72,828||76.8||52.1|
|Old Bexley & Sidcup||64,831||69.8||33.5|
This shows that whilst turnout in the 2019 General Election was fairly close in percentage terms (a spread of 8.9%) the 2021 by-election turnouts vary between 33% and 52% – which is more than twice the General Election variation. General Elections turnouts are usually fairly consistent across the country, but by-elections vary much more, because of both local circumstances and what is happening nationally at the time. The one consistent characteristic between the last General Election in 2019 and the recent by-elections is that the number of registered voters will be very similar between one year and the next for any given constituency. Thus, the Registered Electors column is one to keep in mind.
What struck me as I began to look at these figures, was how similar the Conservative vote was in each constituency for 2021. When expressed as a percentage of the Registered Voters the figures looked even closer:
|2021 Conservative Votes||2021 Conservative votes as % of Registered Electors|
|Chesham & Amersham||13,489||18.5%|
|Old Bexley & Sidcup||11,189||17.3%|
These are all within 3.9% of each other. We can look upon these votes as the Core Conservatives who will vote Conservative through thick and thin. These are, if you like, the tribal voters. One point worth noting is that the core appears to have declined over time from June (Chesham and Amersham) to December (North Shropshire), perhaps as a response to continued demonstrations of chaos within government. Nevertheless, as a rough guide, we can argue that Conservative core voters represent about 15% to 18% of the potential electorate in these three constituencies.
Whilst this is the core, there is another other group of Conservative voters who represent what I call the sceptical conservatives. These are the people, who are ‘small c’ conservatives – centre right, not members of the party, broadly socially liberal, economically right-wing, often private sector, small business and so on. This group is the most interesting in this context, because these are the ones who in 2019 voted for Boris Johnson to get Brexit done, to end the uncertainty and get the economy sorted out. It was this group which gave the Conservatives an 80-seat majority in 2019. But in the three 2021 by-elections, this group withheld their votes:
|2019 Conservative Vote (a)||2021 Conservative Vote (b)||Sceptical conservatives (a – b)||% of Registered Electorate|
|Chesham & Amersham||30,850||13,489||17,361||23.8%|
|Old Bexley & Sidcup||29,786||11,189||18,597||28.7%|
Once again, the percentage of the Registered Electorate that these voters represent, are very similar between each of the three constituencies. With reasonable confidence, we can say that the sceptical conservatives comprise about a quarter of the total potential vote.
The unifying characteristic of these three by-elections was the severely reduced turnout when compared to the 2019 General Election. Whilst a turnout reduction can happen with most by-elections, it tends to happen because of a general discontent and apathy across all parties. The scale of the 2021 by-election reduction is discussed for Old Bexley and Sidcup here. In most years, and for most by-elections, the general political scene is often uninspiring. However, in the case of these three by-elections, it is very clear that the disaffection with politics has been centred upon Conservative voters. Having isolated the group of sceptical conservatives in Table 3, we can check to see if the drop in turnout between 2019 and 2021 comprises solely these sceptical conservatives, or whether other groups are also involved in the disaffection:
|Turnout 2019 (a)||Turnout 2021 (b)||Drop in turnout (a – b = c)||Sceptical conservatives (d)||Changes in turnout by other groups (c – d)|
|Chesham & Amersham||55,978||37,954||18,024||17,361||-663|
|Old Bexley & Sidcup||46,145||21,783||24,362||18,597||-5,765|
From this we can see that at Chesham and Amersham in June 2021, a mere 663 people from other groups declined to vote. Which means that the reduction in turnout was almost solely due to the sceptical conservatives abstaining. At Old Bexley and Sidcup, the disaffection spread to other groups – the abstaining sceptical conservatives were augmented by a further 5,765 abstainers. But this changed the following week in North Shropshire, where the number of disaffected voters (Column c) was exceeded by the sceptical conservatives (who withheld their votes from the Conservative Party) by 5,009. This suggests that about 5,009 sceptical conservatives were sufficiently exercised to vote and shift their votes to other parties – the principal recipients being the Liberal Democrats (see explanation below Table 7).
Nevertheless, in all three constituencies, the vast majority of people who abstained in 2021, comprise the sceptical conservatives.
If these numbers of sceptical conservatives are taken and compared with the turnout for 2019 General Election, we get this:
|Sceptical conservatives (a)||2019 Turnout (b)||Sceptical conservatives as % of 2019 turnout ((a/b)*100)|
|Chesham & Amersham||17,361||55,978||31.0%|
|Old Bexley & Sidcup||18,597||46,145||40.3%|
From this, we can see that the sceptical conservatives comprise between 30% and 40% of the total votes cast during the 2019 General Election. This is a huge and very important percentage which generally gets out and votes. It cannot be ignored or dismissed by the Conservative government, because this group is the one which decides which party governs the country.
Having examined some of what happened to the Conservative votes cast in 2019 and 2021, we must now consider what happened to the Liberal Democrat votes.
In Chesham and Amersham, and North Shropshire, the Liberal Democrats scored massive victories. In Old Bexley and Sidcup, their vote was almost negligible, so we can assume that lacklustre campaigning or some other factor intervened. But in the other two constituencies, a large number of new votes arrived from somewhere. Did they come from disaffected sceptical conservatives, or did they come from other parties such as Labour?
To answer this, we must first look at what happened to the Lib Dem votes between 2019 and 2021:
|Lib Dem vote 2019 (a)||Lib Dem vote 2021 (b)||Lib Dem Increase in votes (b – a)|
|Chesham & Amersham||14,627||21,517||+6,890|
|Old Bexley & Sidcup||3,822||647||-3,175|
In Old Bexley and Sidcup, the Lib Dems lost heavily to the massively reduced turnout in the same way as all the other parties. But their victories in the other two constituencies were generated by very large increases. Given the background of reduced turnout in both of these constituencies, those votes had to come from somewhere.
We now turn to the Labour vote:
|Labour vote in 2019 (a)||Labour vote in 2021 (b)||Labour vote losses (b – a)|
|Chesham & Amersham||7,166||622||-6,544|
|Old Bexley & Sidcup||10,834||6,711||-4,123|
Firstly, let us ignore Old Bexley and Sidcup, because Labour lost to the reduced turnout in the same way as everyone else. But in Chesham and Amersham we see that the Lib Dem gains are almost exactly matched by the Labour losses (Lib Dem gain of 6,890 against Labour losses of 6,544, a difference of only 346). In North Shropshire, the Lib Dem gains exceed the Labour losses by 3,505 (12,314 – 8,809). Here the Lib Dems clearly won new votes from somewhere other than the Labour Party. Let us call these the Lib Dem surplusprotests. The general message is that in both these constituencies, there was most likely a good deal of tactical voting in order to get the Tories out. This tactic clearly worked well.
An argument frequently advanced by a mid-term government which has just lost heavily in a by-election, is that it was some sort of protest vote (“the voters didn’t like our policy on home-grown cauliflowers…”) or some such lame and trivial excuse. The Liberal Democrats are the traditional home for these protest voters. Doubtless their policy on home-grown cauliflowers is more acceptable….
But in the case of these three by-elections, the protest vote is of much greater interest. This is because the three constituencies are fairly uniform in electoral make-up, have produced remarkably similar but dramatic results, and can be viewed over time.
On the one hand, we have the negative protest of the voters who refuse to turn out to support a government that they no longer believe in. In addition to these passive sceptical conservatives, we have another group of voters who turned out in these by-elections, and who were determined to give the government a bloody nose. Many of them appear to be traditional Labour supporters who, in the case of two of these by-elections, have lent their votes to the Lib Dems. This group is left of centre in outlook and appears to be happy to use any means at its disposal to oust an unpopular Tory government. Lending their votes to the Liberal Democrats is no real hardship for them, especially if it is in a by-election.
Other normally non-Labour centrist voters may also be inclined to lend their votes to the Lib Dems with an anti-government objective in mind. This seems to have happened in North Shropshire. Let us call these votes active protests.
Alternatively, they can vote for a new party which has a sensible manifesto, and which is likely to be Euro-sceptic centre-right. There are a number of these new parties which have sprung up in the last two to four years, some of which comprise elements of the old UKIP which split up after the referendum. There is also the rump of UKIP itself, which still survives. These active protesters are the ones who are angry enough to vote, and not just resigned to putting up with serial governmental incompetence by abstaining.
If we put the various active protest votes in each constituency, excluding the transfers from Labour, but now presented in chronological order, we get the following:
|Chesham & Amersham (17th June)||Old Bexley & Sidcup (2nd December)||North Shropshire (16th December)|
|Lib Dem Surplus Protests||3,505|
|Active Protests as % of Turnout:||1.1%||8.0%||15.1%|
In this part of the analysis, although Labour and Liberal Democrat voters are also seeking to achieve change, their votes for their own usual parties do not provide an indicator of the the level of discontent amongst the usual supporters of the Conservative government within the electorate.
Table 8 above shows the apparent growth of the active protest votes over time. But these are not the only kind of verdict upon government performance. We also have those who have withheld their votes -the sceptical conservatives. If we add the sceptical conservatives to the active protest votes, we get the following:
|Chesham & Amersham (17th June)||Old Bexley & Sidcup (2nd December)||North Shropshire (16th December)|
|Active Protests as % of Turnout:||1.1%||8.0%||15.1%|
|Passive Protests as % of Turnout:||31.0%||40.3%||41.4%|
|Total, all protests:||32.0%||48.3%||56.5%|
Here we can see that the total electoral discontent has risen from 32% to 56% in a mere six months. Although the passive protests are defined by the fact that they did not vote in the by-elections, it is vital to stress that this dramatic rise in protests against the government is amongst those who would normally vote in a General Election for the Conservatives. Furthermore, it is sitting there, latent and waiting for a focus to engage it.
- Table 1 – Percentage turnouts are more consistent between constituencies in a General Election than they are in a by-election.
- Table 2 – Conservative votes in the by-elections were very consistent as percentages of the Registered Voters. This is the Conservative core and is about 15% – 18%.
- Table 3 – The Conservative voters in the General Election who either abstained or did not vote Conservative in the by-elections are classified as sceptical conservatives. These are also a consistent percentage and are about 25% of the Registered Voters.
- Table 4 – (i) At Chesham and Amersham, the disaffected voters were nearly all sceptical conservatives. (ii) At Old Bexley and Sidcup, the disaffection was widespread and all main parties lost votes. (iii) In North Shropshire, about 5000 sceptical conservatives shifted their votes to other parties. (iv) But in all three places, the vast majority of people who abstained were sceptical conservatives.
- Table 5 – The proportion of sceptical conservatives in each of these by-elections increased over time from 31% at Chesham and Amersham to 41% at North Shropshire.
- Table 6 – (i) At Old Bexley and Sidcup, the Lib Dems lost heavily, as did the other parties. (ii) At Chesham and Amersham Lib Dem gained 6,890 votes over their 2019 performance. (iii) At North Shropshire, they gained 12,314 additional votes.
- Table 7 – (i) at Chesham and Amersham, it looks as if Labour supplied all the additional votes that caused the Lib Dem victory. (ii) At Old Bexley and Sidcup, Labour lost votes along with everyone else. (iii) At North Shropshire, Labour’s losses do not explain the totality of Lib Dem gains. (iv) The Lib Dems gained an additional 3,505 votes from sources other than Labour. These are classified as Lib Dem Surplus votes.
- Table 8 – The active protest votes (including the lib Dem surplus) increased over time for 1.1% to 15.1%.
- Table 9 – When the above are added to the sceptical conservative votes, the protest increased from 32% to 56.5% over time.
This analysis shows that the scale of voter discontent with Boris Johnson’s government is very real, serious, and is growing rapidly. The discontent is not just expressed amongst the usual opposition parties and their regular voters, but also amongst the group who normally vote Conservative. Not only is this group of ‘small c’ conservative protest voters growing, but it now represents well over half the people who normally come out to vote. This is a shocking but unavoidable conclusion.
Boris Johnson has effectively lost the electoral base which gave him his enormous win in 2019. This electoral base was happy to lend their votes to the Conservatives, provided Brexit was done – and the economy put back on track with a concomitant restoration of economic certainty.
When Covid intervened, the governance of the country descended into chaos. At first, this was quite slow, but now that descent has reached the speed of a black run on a ski slope. Or perhaps worse – Boris now resembles Eddie the Eagle as he hits the bottom of the ski jump and is flipped into the air, arms, legs and skis flailing. The landing will not be pretty.
There is no question that the Conservative Party is now preparing to ditch Boris. When it finally happens, they will do so with their customary ruthlessness. I suspect the end of January 2022 will be a good time. The party procedures may drag this timing out a little, but the skids will be pushed under him when Parliament returns in the New Year.
What happens then is very much in the lap of the gods of the Conservative Party. If a ‘caretaker’ leader is appointed, a Remainer perhaps, who is deemed steady but worthy, then the country will continue in the chaotic slipstream which is Boris’s legacy. The electoral revolution which has been brewing for months and is manifested within this analysis, will continue to grow. If the government does not stand up to the lockdown hawks on SAGE, for example (preferably by sacking the lot of them), then the protests will become increasingly vocal, street-borne, and violent.
The electoral consequences of this scenario are the possibility that every Conservative seat in the country is likely to change hands. Especially vulnerable, in my view, are the so-called ‘safe’ Tory seats (just like the ones examined in this analysis) encumbered with a boring, steady-as-she-goes, safe-pair-of-hands, old-school, patrician Tory MP. This type has done little of anything, has voted consistently with the government and is hoping to be rewarded for their unimaginative but loyal tenure with at least a Knighthood and preferably a seat in the Lords. If I am right, vast swathes of the Home Counties will change hands.
The rest of the country, outside London and the Home Counties, tends to elect a Conservative MP as a probationer – one whose performance is constantly under review. It is unlikely that Northerners or Midlanders will tolerate an MP who has not done his or her utmost to defend their constituents against the excesses of an increasingly authoritarian government. Here in Cornwall, traditionally a Liberal stronghold, it has taken a decade or more to turn it Tory blue. I suspect it will revert to Liberal Democrat yellow again fairly soon.
Another possibility is that the Parliamentary Conservative Party recovers some common sense from somewhere and elects a leader who sorts the mess out: stops the cross-channel illegal immigration, deports the illegal immigrants, throws out the entire SAGE committee and their recommendations, gets children back to school, stops any hint of mandatory vaccination – and thereby stops and even reverses the exodus from care home staff and NHS nurses, re-establishes ethics within the medical professions whereby the individual patient is treated, sorts out the NHS impending disaster of untreated cancers, sorts out the border problems in Northern Ireland, trounces once and for all the incompetent Sturgeon and restores faith in the Union….and so the list of desperately urgent problems goes on and on.
In this scenario, the Tories might, just, recover their credibility with the electorate and survive as a political force, perhaps even retaining their hold upon government.
But there is another group that needs to be discussed: the collection of other centre-right parties who are now forming a small but vocal opposition to the more authoritarian policies of the current government – Reform UK run by Richard Tice, The Reclaim Party run by Laurence Fox, UKIP run by Neil Hamilton, and the Heritage Party run by David Kurten. All of these parties have roughly similar manifestos. Their current voter support is also similar in outlook.
Unfortunately, these parties appear to have their differences between each other, and so their vote is split and their effect upon the political landscape is often diminished to the point of insignificance. And yet, there they are – very active, and with an enthusiastic membership. They cannot be dismissed because, as seen in Table 8 above, their influence appears to be growing. But at the moment, they present the average voter with a negative perception: several leaflets arriving through their door from four different parties, all apparently saying much the same thing. This only serves to confuse and maybe even irritate the voter, who will view a vote upon any one of them as a wasted vote – because it won’t be effective. The voter, who correctly regards their vote as precious and which should be used as effectively as possible, might just as well vote for the Monster Raving Loony Party for all the political good it does. None of us wish to place our bets upon an apparently lame horse. This perception must be changed if any of these parties are to succeed.
The only solution to this problem is for the parties to combine their campaigns, perhaps informally, on a constituency-by-constituency basis. This then concentrates the minds of the voter upon one non-mainstream party candidate.
It is perfectly possible for this to happen. During the EU referendum, this is precisely what happened in Cornwall, with UKIP helping out the Vote Leave campaign and Tories, who helped out others. The prize for doing this informal co-operation lies in the vast numbers of voters (32% – 56%, see Table 9 above) who are currently disenfranchised by the appalling ineptitude of both government and Opposition alike.
The prize of millions of votes is waiting for them. All they have to do is act as a team and grasp it.
At the end of my analysis of the Old Bexley and Sidcup by-election last week, I quoted the comment attributed to Admiral Yamamoto after the Japanese attack upon Pearl Harbor. I did this because, after completing that analysis, I realised that the potential electoral rebellion against Boris Johnson was large. Only on completion of this analysis have I come to realise that the sleeping giant is not large.
He is huge.
Thank you for this analysis and discussion. It is nice to see data driven commentary. The Tories need to change today’s narrative. Changing the leader is always one step. But the ‘levelling up’ narrative albeit hurt badly by the decision not to build a new rail line between Leeds and Manchester is still an important part of the mix. If you have a chance take a look at my new co-authored book Reclaiming Populism: How Economic Fairness Can Win Back Disenchanted Voters (Polity Books). It is data rich analysis. There is a way forward. Happy to send you a copy.
Thank you Paul. Sorry to have taken so long to respond. The book sounds very interesting and touches upon areas which have bothered me for some time.