From Red Walls to Red Bridges: Politics After Class

The two polling reports and excellent analysis by Peter Kellner provide the Labour Party with a clear roadmap for winning.

Coming close on the heels of the UK’s local elections, the Macron victory in France and the re-emergence of the centre ground in Europe, they show exactly what the problem is and the solution. The loss of so many Conservative council seats last week and Labour’s return to second place in Scotland make clear the appetite among UK voters for progressive politics.

Peter’s previous report, FromRed Walls to Red Bridges: Rebuilding Labour’s Voter Coalition, showed that the “working class” as traditionally defined by occupation is now a minority of the population. A Labour strategy based on a narrow class appeal would be doomed.

This second report shows that the long-term trend of people thinking class is less relevant continues.

However, interestingly, a majority still think of themselves as working class. When we probe, we find that this is based not on the old objective notions of class according to a person’s place in the processes of production, but rather on subjective self-identification about how they see themselves. So, you have crossover groups, quite large, of traditionally defined working-class people who believe they’re middle class, and vice versa.

This majority’s definition of working-class jobs covers a broad range of occupations that would not traditionally be categorised as such. For instance, around a third of people regard a nurse as equally likely to be in either the working-class or middle-class categories.

And it is as much about values as traditional notions of class. The values that shaped you when you were growing up. Better values are associated with the working class, with the middle class deemed more “greedy” and the working class more “honest”.

This class “crossover” effect then yields some fascinating results when combined with how people voted on Brexit.

The polling splits people into four groups. Two believe they’re working class, though one of these is by traditional definitions middle class; two believe they’re middle class, though one is traditionally defined as working class.

The first two “working-class” groups comprise 62 per cent of the country, equally split. But they couldn’t be more different on Brexit.

The first group, the “real” working class, are likely the traditional Labour vote of the north – constituencies like my old one of Sedgefield. There is a residual loyalty vote for Labour, declining but still present. People in this group will be culturally more conservative, and possibly older. Labour has a cultural problem with them.

They voted 70-30 per cent Leave, but the Tories only had an 11 per cent lead among them in 2019.

People in the second group are more likely university educated, possibly working in the public sector, living in the cities, okay with modern culture, but see themselves as opposed to the Tory establishment and privileged class.  

This 31 per cent voted 56-44 per cent Remain, and Labour had a 5 per cent lead among them in 2019.

This is what I would expect.

The interesting thing is the other 38 per cent.

These two groups believe they’re middle class, though only the first truly is as traditionally defined.

The first group comprises 26 per cent of the population, are likely professionals, in the private sector, living in the commuter belt, and are worried about issues like tax and economic competence.

The second is your aspirational working class and represents 12 per cent of the country. These are the people who believe they’re on the way up, and don’t want government getting in their way.

The 26 per cent – the traditional middle class – voted Remain 57-43 per cent, but the Tories had an election lead of 22 per cent.

The 12 per cent – blue collar but regarding themselves as middle class – voted 53-47 per cent Leave, but the Tory lead among them was a whopping 32 per cent.

Labour clearly has a problem with parts of the traditional working class. But the bigger problem lies with these latter two groups: the middle class and the aspirational working class. Yes, the traditional blue-collar voters gave the Conservatives a lead, but that was much less than the 40 per cent difference on the Brexit vote.

But when it comes to those who identify as white collar, the disparity between the Brexit vote and the Conservative vote is huge. By large margins, these voters preferred the Tories even though on Brexit there was a small majority for Remain.

The only sensible explanation for this was their fear of Corbyn’s Labour in 2019. A large number voted Conservative despite disagreeing with the party on Brexit. They thought Labour’s far-left economic policy was a bigger threat than Brexit.

The obvious conclusion is that Labour must cure both its cultural and its economic-aspiration problems. It must do the first to regain those voters who went Tory despite being traditional blue collar. These are the northern “Red Wallers”.

But Labour must also do the second to have any realistic chance of winning.

Of course, this is entirely familiar. An alliance between the middle class and the aspirational working class has always been key to winning. And as traditional class has declined in relevance and identity and has become more fragmented, it becomes even more important.

The challenge for Labour is to build a shared understanding of what its problem is. A lot of Labour-policy debate focuses on what makes Labour feel good about itself. The “betrayal” thesis from the left casts an ever-present shadow. So, one part of Labour wants to make spending commitments to show the party that it cares about issues; talks constantly about the poorest in our society because, of course, it cares, but has less to say about those a bit better off; and eagerly takes an opportunity to lambast business for greed, but treads warily in congratulating business for wealth creation.

The point is not that Labour in government should not spend more on the NHS, alleviate poverty or introduce protections against business malpractice. Of course it should. As indeed, contrary to far-left criticism, the last Labour government did – with record spending on health and education, big reductions in child, family and pensioner poverty, and the minimum wage for workers.

The point is that the electorate doesn’t doubt Labour stands for these things. It doesn’t dispute Labour’s commitment to social justice. Parts of the Labour Party do, but the broad public? No, that’s not their concern about Labour. It’s Labour’s concern about itself.

Their worry about Labour is not whether Labour has principles. They accept that the party does. That is, in part, their anxiety. That in pursuit of these principles, Labour might put those principles before them and their families.

The difference in the psychology of the public towards the Tories and Labour is important.

They know the Tories care only about power. But whereas your Labour activist sees this as evidence of Tory iniquity, for the public, it’s curiously comforting. It makes them think that the Tories are always going to be trying to please them. So, Boris Johnson may be unscrupulous and only interested in self-preservation, and the electorate may feel he doesn’t deserve to carry on governing, but that doesn’t mean they won’t carry on voting for him – unless his survival strategy becomes a risk for them, and Labour is regarded as the safe alternative.

Labour has made huge strides forward under Keir Starmer’s leadership. His challenge after the Corbyn years has been immense; but the changes have also been immense.

The country is looking for an alternative to the present Tory government, and there is a growing feeling it should be put out.

Labour’s task is now to convince voters that Labour can govern; that it has a programme that will move the country forward; that it can be radical without being dangerous.

Fortunately, Labour can do just that because, again sensibly, Keir Starmer focused first on sorting out the Labour Party so that he has control of it, and now can turn his attention to formulating the policy agenda.

That agenda consists of three things in my view:

• Staking out a position on the “culture-wars” issues that plants Labour’s feet clearly near the centre of gravity of the British people, who want fair treatment for all and an end to prejudice, but distrust and dislike the “cancel culture”, “woke” mentality.

• An economic policy which – as Rachel Reeves is doing – moves Labour from a wish-list to a credible framework that supports business and makes people comfortable with the prospect of Labour running the economy.

• A plan for the country’s future, from health and education to building the nation’s infrastructure – digital and physical – and for long-term energy security, which makes an incoherent governing strategy based on the personal survival of Boris Johnson a risk.

And it is a risk. The absence of long-term strategy means the underlying problems of the country are unaddressed. Difficult decisions get shelved not because they’re wrong but because they’re difficult.

The core challenge for Britain is our poor growth and productivity. We’re looking at future trend growth rates of under 2 per cent, way below those of the first years of the century. The productivity record of the past 15 years is even worse. GDP and productivity figures mean nothing to normal people. But the reality is they translate into the difference between funding public services well or not, and higher real wages or not.

There is no sustainable way out of the “cost of living crisis” without them.

Across the board, we need radical reform not engineered by old ideology, either left or right. Brexit – which should have been the signal for such far-reaching reform – has instead resulted in lost trade, reduced growth, a crisis in Northern Ireland, more bureaucracy, border chaos and massively increased hassle in travelling within Europe for British passport holders.

We have an enormous climate-change ambition but, as backing-off planning reform shows, we lack the strategy to fulfil this ambition while also keeping energy costs low.

And we need a plan to harness the enormous potential of the technology revolution, without which inequality and low growth will become even more entrenched.

Such a programme will be fiercely contested by the laissez-faire right and those on the left who want a series of dividing lines with the Tories that may have resonated in the 1980s but have no place in the third decade of the 21st century.

Naturally, Labour should continue to point to the unfairness of our present society and to out-of-touch Tories who make the rules but don’t play by them. However, the public, though they will agree with much of what Labour says, will not vote Labour come an election unless they can be sure that Labour understands the modern world, is not driven by ideology but by a coherent analysis of what the country needs, and will offer something radical that is not risky.

The Tory campaign at the next election will be utterly predictable. They will say: “Labour has big spending plans it can’t afford; an addiction to woke; and Corbynism is not dead in Labour, they’re still hanging around. So, you may not like us, but the alternative is a risk you can’t take.”

Because predictable, it can be countered.

Labour’s policy programme should reassure the Red Wallers and the middle class on cultural issues and the economy.

However, reassurance is necessary but insufficient.

The plan should also show that Labour is prepared to think boldly where the Tories can’t or won’t.  

Let me give two examples. But they’re merely illustrative.

Immigration. The purpose of the Rwanda ploy is not to solve the problem of illegal immigration. It is to provoke outrage and therefore establish in the public mind that the Tories are “tough” on immigration and Labour “soft”.

However, there is a solution to illegal immigration. It is to use the latest technology to create biometric ID for each citizen in order to access services and to work. At any one time, there are – my Institute estimates – anywhere up to 1.2 million people here in the UK illegally. The numbers coming on boats across the channel are a tiny fraction of this figure.

Embrace the solution. There will be a row. But we will win it.

Second, Europe. Believe me, I get it. People don’t want to revisit the Brexit decision. It has been done. It was hugely divisive. The country wants to put it behind us. Labour is understandably anxious that even mentioning it will turn the Red Wallers blue.

Labour should make clear we’re not rejoining the EU or revisiting the decision.   

But we need a positive working relationship with our continent, and the absence of one is a real problem absolutely identified with the Tories. Remember the promises? No difficulties on EU trade, oven-ready agreement with the United States, no issue with the Northern Ireland border, reductions in bureaucracy and regulation, more public money to spend on the NHS.

You can’t let your political opponents get away with a shocker like this.

My suspicion is that the middle-class 38 per cent will want to punish the Tories for these false promises, provided they’re not terrified of Labour, and that the 31 per cent traditional working class who voted heavily Leave won’t feel the same intensity they felt in 2019, now that the Corbyn millstone has been removed.

The bane of progressive politics is to think the choice is between being voter-friendly and boring, or exciting and voter-repellent. It all comes back to confusing “radical” with old-fashioned leftism.

But it doesn’t need to be like that. And Labour under Keir Starmer has the chance to prove it.   

Tony Blair

Executive Chairman

For more than 120 years, the disputes at the heart of British politics have concerned social class, prosperity and the links between them. In election after election, Labour was more popular among working-class families while the Conservatives were more popular with middle-class voters. The division was never absolute: the Conservatives’ run of big victories in the 1950s and 1980s would have been impossible without the support of a third of manual workers and their families. When Harold Macmillan addressed a Conservative Party rally in 1957, stating that “most of our people have never had it so good”, he was partly appealing to Benjamin Disraeli’s “angels in marble”.

Even so, Labour retained the loyalty of most working-class voters.

In 2019, the pattern of class voting was reversed. The Conservatives enjoyed a bigger lead among voters traditionally regarded as working class – those classified as C2DE – than among those considered middle class – classified as ABC1. The result was that the Tories gained a range of seats in industrial England. In From Red Walls to Red Bridges: Rebuilding Labour’s Voter Coalition, we showed how age and education had replaced those traditional classifications as the main influences on the way people voted.

However, social class is about more than occupational classifications, which were designed for a different era, dominated by heavy industry – coalmines, shipyards, steelworks, mills and massive factories.

This report looks more broadly at social class in our post-industrial age – how people view class differences, how they think of themselves and the connection between class and politics today. It draws on the results of a two-part Deltapoll survey of close to 4,300 adults across Britain that took place in mid-January and March 2022. The survey uncovers a big discrepancy between the way voters are conventionally classified and the way in which they think of themselves.

The data show that social class still matters to some extent, affecting the way a minority votes. However, its impact has changed and diminished. Few voters think social class matters to them personally in their daily lives. The data also suggest that the relevance of class will continue to decline. Yet the two main political parties still have clear class identities rooted in their rival histories. Both risk being stranded with outdated reputations in an era in which most voters have abandoned the class loyalties of the 20th century.

Polling Highlights

• According to conventional classifications based on occupation, middle-class (ABC1) adults outnumber their working-class (C2DE) counterparts by four-to-three. But when people are asked to assess their own social class, six-in-ten say they are working class.

• The main influence on people’s own class identity is not their occupation, income, qualifications or whether they own their own home. It is the class of their own family when they were growing up.

• These two different ways of analysing class produce different political outcomes. At the 2019 election, Conservatives outnumbered Labour voters by similar amounts among both ABC1 and C2DE voters. But when votes are analysed by people’s self-defined social class, Conservative support is tilted towards those who say they are middle class while Labour remains relatively stronger among those identifying as working class.

• The most pro-Conservative voters at the 2019 election were manual (C2DE) workers who regard themselves as middle class. The Tory lead in this “crossover” group was greater than among the more traditional white-collar (ABC1) workers who consider themselves middle class.

• The most pro-Labour voters also belonged to a crossover group: ABC1 voters who regard themselves as working class. Labour enjoyed a narrow lead among these voters while lagging some way behind the Conservatives among C2DE, self-defined working-class voters.

• Many white-collar jobs, classified as ABC1 and hence middle class, are widely seen by voters today as working-class occupations. This helps to explain why far more people describe themselves as working class than are identified in conventional survey research.

• Around half the public think social class makes little or no difference to their character. But among the other half, who do detect a difference, most regard friendliness and honesty as working-class traits while greed is thought to be far more common in the middle classes.

• Two-thirds of voters believe that Britain is still beset by a big division between the lives of working-class and middle-class families. And the Labour and Conservative brands are still widely associated with those divisions.

• However, while most people say social class mattered to their parents’ generation, only one-in-five say it matters to them personally.

• When people are asked about the factors that influence their identity, only one-in-four say that either social class or political outlook matter much. Their family roots, nationality (British and/or English/Welsh/Scottish) and hobbies are bigger considerations for many more people.

When polling company Deltapoll asked respondents which class they belong to, large numbers chose a different class to the one allocated to them by the conventional social-grading system. This conventional classification of voters, used by survey and polling companies, is derived from the occupation of the head of each household (or, for those who have retired, the job they had before retiring):

Middle class or white collar or non-manual

A: Professionals, very senior business managers and top-level civil servants.

B: Middle management in business, local government and the civil service, and top management or owners of small businesses, education and service establishments.

C1: Junior management, owners of small establishments and other non-manual workers.

Working class or blue collar or manual

C2: Manual workers who are skilled and/or are responsible for other people.

D: Semi-skilled and unskilled manual workers, and trainees or apprentices of skilled workers.

E: Long-term recipients of state benefits, unemployed or those off sick for more than six months, casual workers without a regular income and intermittent workers receiving income support.

Figure 1 compares the answers that respondents gave when asked to describe themselves as upper, middle or working class with their conventional classification based on the occupational categories above.

As can be seen, four-in-ten (42 per cent) AB voters regard themselves as working class while three-in-ten (29 per cent) DE voters say they are middle class. In the case of C1 voters, almost two-in-three (64 per cent) describe themselves as working class despite belonging to the broad group classified as middle class.

Overall, 62 per cent say they are working class – compared with the 43 per cent who are actually classified by the conventional grading system as the equivalent C2DE.

If the link between someone’s occupation and their own class identity is so weak, what factors do explain whether they think of themselves as middle or working class? One set of figures, which stands out as shaping people’s class, sheds some light.

As well as asking people to which class they personally belonged, Deltapoll asked: “If you had to say, which class would you say your parents belonged to when you were growing up?” Leaving to one side the small minority who replied “don’t know” to either question, this is how British voters see themselves:

Voters who say they belong to the same class as their parents:

Middle or upper class: 26%

Working class: 58%

Total same class: 84%

Voters who say they belong to a different class from their parents:

Voter says working class but identifies parents as middle or upper class: 4%

Voter says middle or upper class but identifies parents as working class: 12%

Total different class: 16%

As shown by the percentages above, people’s sense of their own class is overwhelmingly rooted in the circumstances in which they grew up. Where people say they belong to a different class, they are far more likely to have moved from a working-class childhood to the middle class today, but they still only represent 12 per cent of Britain’s total population. What is clear is that other features widely associated with class – occupation, income, qualifications and housing tenure – have a far weaker correlation to the way people define themselves than their family roots. Later in this report, we return to the impact of people’s upbringing on their outlook.

Turning to the relationship between class identity and the way people vote, Figure 2 reveals how the contrast between the two ways of assessing class makes a significant difference. Deltapoll asked respondents how they had voted in the 2019 general election and how they would vote today (as of mid-January 2022).

The figures for 2019 tell much the same story as other surveys: big Conservative leads among both ABC1 and C2DE voters, with – for the first time – the Conservative lead among C2DE or working-class voters (16 per cent) overtaking that of ABC1 or middle-class voters (8 per cent). However, when we analyse the figures according to the way that voters describe themselves, a very different picture emerges. The Conservatives enjoyed a two-to-one lead over Labour (49 per cent over 25 per cent) among voters who describe themselves as middle class while the two parties are almost neck-and-neck among working-class voters.

It would seem from Figure 2 that the conventional ABC1/C2DE division of the electorate has lost its value as a way of examining voting behaviour and that we should concentrate instead on the way people view themselves. However, that does not always seem to be the case.

Consider the Brexit referendum in 2016, when:

ABC1 or middle-class voters divided 57 per cent over 43 per cent in favour of staying in the EU (representing a Remain lead of 14 per cent) while C2DE voters divided 64 per cent over 36 per cent in favour of Brexit (representing a Leave lead of 28 per cent).

However, when considering self-defined class, middle-class voters divided 55 per cent over 45 per cent in favour of staying in the EU (representing a Remain lead of 10 per cent) while working-class voters divided 57 per cent over 43 per cent in favour of Brexit (representing a Leave lead of 14 per cent).

The difference in the percentage-point gap between the class difference when using these two definitions changed from 42 based on the conventional classification to just 24 based on the self-defined. In other words, the class divide on Brexit was greater when analysed in terms of the conventional ABC1/C2DE classification than self-defined social class. There seems to be no consistency in the application of different ways of assessing class to the way people vote.

What is going on? History provides some clues.

Karl Marx developed the concept of class in an era when a small minority owned wealth in the form of land or capital, or both, while virtually everyone else scratched a meagre living as their employees or tenants. As industry and the wider economy evolved, an employed but better-educated and better-paid, white-collar middle class emerged.

In the early years of the 20th century, the infant Labour Party saw both kinds of worker as its natural allies. In 1918, its new constitution justified its commitment to common ownership by declaring its ambition “to secure for the workers by hand or by brain the full fruits of their industry”. The aim of an alliance between manual and non-manual workers could not be clearer.

However, as the economic changes of the 20th century unfolded, that aim was thwarted. In the years following the second world war, the market-research industry developed the ABC1/C2DE-class system outlined above.^[_] It was quickly adopted for political research and proved to be extremely effective at distinguishing blue-collar (“workers by hand”) from white-collar (“by brain”) political loyalties. The former were mainly Labour (apart from Disraeli’s “angels in marble”), the latter mainly Conservative.

The terms non-manual, white-collar and middle class were used interchangeably, as were manual, blue-collar and working class. This simple arrangement was not disturbed by the fact that academic studies, from time to time, showed many people defining their own class differently from the ABC1 classification. The big picture worked, whichever definition was used.

This is no longer the case. The divergence between conventionally classified and self-defined middle-class and working-class voting patterns is plainly significant. One element of this is that Labour’s 1918 concept of class is rivalling, and may be overtaking, the market researchers’ concept.

While blue-collar and white-collar workers divide their party loyalties in much the same way as each other, they view Brexit differently, so the challenge is not to discard the ABC1/C2DE system altogether, but to examine both together, and how they interact.

As we shall see in the next section, the general public regards a number of ABC1 occupations as working class even if they are non-manual. But Deltapoll’s data also tell us something that goes well beyond an academic curiosity: they have important lessons for politics and elections in the years ahead.

This can be seen by dividing the electorate into four broad groups that draw on both definitions of working and middle class – and which consider how they voted at the last election as well as the Brexit referendum. They are ranked below from the most Conservative to the least:

Blue-collar aspirational are conventionally classified as C2DE, but regard themselves as middle class. They comprise 12 per cent of the electorate. They are the “upmarket angels” who voted heavily for the Conservatives at the last election: 54 per cent for the Conservatives over 22 per cent for Labour, representing a Tory lead of 32 per cent. While they are far more Conservative than voters as a whole, their stance at the Brexit referendum fell in line with the national average: 53 per cent for Leave over 47 per cent for Remain.

White-collar conventional think of themselves as middle class and so do the market researchers who classify them as ABC1. They comprise 26 per cent of the electorate. In 2019, they voted 50 per cent for the Conservatives over 28 per cent for Labour, representing a Tory lead of 22 per cent. In the Brexit referendum, however, they opted for Remain over Leave by 57 per cent over 43 per cent.

Blue-collar traditional define themselves as working class, in line with their C2DE classification. They comprise 31 per cent of the electorate. In 2019, they voted 49 per cent for the Conservatives over 38 per cent for Labour, representing a Tory lead of 11 per cent. Despite their election vote dividing in line with the national average (which produced a 12 per cent lead for the Tories), this group voted heavily for Brexit by 70 per cent over 30 per cent.

White-collar progressive regard themselves as working-class despite having non-manual, ABC1 occupations. Comprising another 31 per cent of the electorate, this is the only one of the four groups that preferred Labour in 2019, their vote of 44 per cent over 39 per cent for the Conservatives, delivering a Labour lead of 5 per cent. As for Brexit, this group divided 56 per cent over 44 per cent in favour of Remain.

As shown, the real drama in the 2019 election lay with the responses of the “crossover” groups whose self-defined class differs from the one allocated to them by market researchers. Blue-collar aspirational – in other words, families with incomes from manual work but who describe themselves as middle class – overwhelmingly voted Conservative. In the opposite corner of Britain’s demographic map, Labour enjoyed a modest lead among the white-collar progressives – in other words, families with incomes from non-manual work but who regard themselves as working-class.

In contrast, the “consistent” groups of white-collar conventional and blue-collar traditional were far less polarised. They rank second and third, with Conservative leads of 22 per cent (ABC1/middle class) and 11 per cent (C2DE/working class).

This is curious. If there is a class element to the way people vote, one might expect those who are “objectively” and “subjectively” similar to have their political loyalties reinforced, with the Conservatives doing best among those who are both ABC1 and self-defined middle class, and Labour doing best among those who are both C2DE and self-defined working class.

Instead, it is the “crossover” groups, whose mixed class identities might lead them to congregate around the national average in terms of voting loyalties, who actually live furthest from that average.

In particular, the Conservatives’ astonishing success in dominating the votes of the blue-collar aspirational helps to explain their national victory – and their gains of the “red-wall” seats dominated by C2DE voters. These are people who used to vote Labour en masse but by 2019 saw themselves as middle class and voted Tory. Altogether, around 2 million self-defined middle-class members of C2DE grades voted Conservative.

Of course, the Conservatives also did well among the larger number of blue-collar traditional voters who maintained their working-class identity. These are the more traditional “angels in marble”. Around 4 million of them voted Conservative. If anything, the curious thing is why the Tories did not do even better with this group. The gulf between their Conservative vote and their support for Brexit was far larger than in any of the other three groups:

Blue-collar aspirational:

• C2DE/middle class

• Leave: 53%

• Conservatives: 54%

• = 1-point difference

White-collar conventional:

• ABC1/middle class

• Leave: 43%

• Conservatives: 50%

• = 7-point difference

Blue-collar traditional:

• C2DE/working class

• Leave: 70%

• Conservatives: 49%

• = 21-point difference

White-collar progressive:

• ABC1/working class

• Leave: 44%

• Conservatives: 39%

• = 5-point difference

It seems from these figures that Brexit may have helped the Conservatives triumph among aspirational blue-collar workers who regard themselves as middle class, but it had less effect (though still some) among more traditional blue-collar workers who still identify as working class. Put another way, the rebellion of traditional working-class voters against the established order was more decisive in the Brexit referendum than in the 2019 general election. Without that supercharged blue-collar determination to “take back control”, the UK would still be part of the European Union.

How would the same groups vote today in a general election? These are the responses to Deltapoll in January 2022:

Blue-collar aspirational: A Conservative lead of 7 per cent among this group, with a swing to Labour of 12.5 per cent.

White-collar conventional: A Conservative lead of 2 per cent, with a swing to Labour of 10 per cent.

Blue-collar traditional: A Labour lead of 15 per cent, with a swing to Labour of 13 per cent.

White-collar progressive: A Labour lead of 21 per cent, with a swing to Labour of 8.5 per cent.

From an overall shift in 2019 of 12 per cent (45 per cent over 33 per cent) in favour of the Conservatives to a 9 per cent swing (41 per cent over 32 per cent) in favour of Labour according to the January 2022 survey, it is not surprising that Conservative support has fallen in each of the four groups. Although there are some differences in the swing movements of the four groups, the underlying structure of the electorate has still not changed: the “crossover” groups continue to be more polarised than the “consistent” groups.

To investigate the links between class and voting further, Deltapoll asked people how they viewed the political parties. To a large extent, voters still view the contest between Labour and the Conservatives as a class battle. Considering each of the other main parties, more than six out of ten respondents say they represent both the middle and working classes equally, or say “neither” or “don’t know”. In contrast, almost two-in-three take sides when it comes to Labour and the Conservatives. Among them, huge majorities say the Tories represent the interests of the middle class while Labour represents the interests of the working class.

Some differences emerge when we analyse responses to this question by both social class and party allegiance. By far, the biggest differences concern the way in which Labour voters perceive their own party. By a margin of 68 per cent over 7 per cent, working-class Labour voters think Labour represents the working class rather than the middle class while the figures for middle-class Labour voters on this subject are much closer at 44 per cent over 21 per cent. This indicates that a significant minority of today’s middle-class Labour voters think their party has discarded its working-class roots.

There is no sign of any such development among Conservative voters. Whether they see themselves as working or middle class, very few think of the Tories as a working-class party. It retains its support among the original “angels in marble” despite the fact that it is still not seen as a working-class party today.

What lies behind the contrast between the conventional classification of voters and the way they see themselves? Deltapoll listed 19 occupations and asked people whether someone doing that job was more likely to be working or middle class. In some cases, voters have a radically different view from market researchers.

Few people think any of the 12 conventional C2DE occupations are middle class. However, the pattern is different for the seven ABC1 occupations listed. In four cases, more respondents think they are more likely to be done by working-class rather than middle-class workers. And with only one of the seven – a doctor at a local GP surgery – does a clear majority associate the profession with the middle class.

The figures above relate to the public as a whole. There are also significant differences between the responses of working-class and middle-class voters, as shown in Figure 6.

Apart from local GPs, where the views of working- and middle-class respondents are much the same, working-class voters are more likely than middle-class ones to associate each occupation with the working class – and vice versa for middle-class respondents.

Consider nursing: conventionally classified as a skilled manual and, therefore, C2 occupation. By more than four-to-one (57 per cent over 13 per cent), working-class voters agree that this is a working-class job. However, middle-class voters divide fairly evenly (31 per cent over 28 per cent). Similar patterns can be seen with owners of local takeaways, five-star hotel receptionists and office secretaries. Working-class voters divide evenly on the class of teachers in state schools while most middle-class voters view this job as a middle-class one.

These figures help to explain why Britain divides 57 per cent ABC1 and 43 per cent C2DE – but 62 per cent working and middle class combined versus 38 per cent upper class. Millions of people simply put the dividing line between the middle and working class in a different place from market researchers. It’s why people in C1 occupations – junior management and other non-manual positions – are far more likely (by two-to-one) to regard themselves as working class rather than middle class.

More generally, we are seeing the consequences of the huge change in Britain’s economic and social structures over recent decades. The class division used to be clear-cut between people doing heavy manual work in big factories, coalmines, mills, shipyards and so on – and those working in offices or other non-manual workplaces. The terms blue collar and white collar were widely used as alternatives to working class and middle class, reflecting an important truth about, and clear dividing lines between, the two kinds of employment. And not just employment: mainly middle-class constituencies elected Conservative MPs while working-class communities returned Labour MPs, often with huge majorities.

Today, those dividing lines are more blurred. This does not make the labelling of ABC1 and C2DE people wrong – but it is becoming less useful. It has also become a bad predictor of whether a constituency will send a Labour or Tory MP to Westminster.

While conventional classifications flow from occupational status, most people have a far wider view of what makes someone working class or middle class. As we have seen, family roots are the dominant influence on an individual’s own class identity. However, most voters have a wider notion of the connections between social class and people’s lives.

When people are asked which three or four things from a list of 12 they most associate with being middle class, working in an office comes last – even though it is a crucial part of allocating ABC1 status to millions of people. Two of the top three answers are about prosperity: having an average income and owning a home. Had this research been done 30 years ago, it’s a safe bet that it would have detected “angels in marble” among those people who took advantage of the right-to-buy sale of council homes in the 1980s.

As Figure 7 shows, there are small, but only small, differences in the views of working- and middle-class respondents – apart from the fact that middle-class voters are twice as likely as working-class ones to see “working in an office” as a middle-class trait. However, “working in an office” still places in the bottom half of the table, with a percentage of 19.

Deltapoll also asked respondents to pick three or four from 12 possible working-class traits. As far as possible, these were mirror images of the middle-class traits.

The most striking difference with the middle-class results is that “having a manual job” comes top in Figure 8 whereas “working in an office” comes last in Figure 7. In a way, this is not surprising because Figure 6 showed that the most obvious manual jobs were cited by large majorities as working class. Even so, it is striking that fewer than half of all respondents pick “having a manual job” as one of their indicators of working-class life.

This time, there are bigger differences between views on opposite sides of the class divide. By double-digit margins, working-class voters are more likely to cite “a manual job”, “an average or below-average income” and “growing up in a working-class family” as key working-class traits – these are factors that loom large in someone’s identity and daily life. In contrast, middle-class voters are more likely than their working-class counterparts to cite reading a tabloid paper or “belonging to a blue-collar trade union”. These factors could reflect attitudes of those people who view working-class life from the outside.

The broader point is that class, as viewed by voters rather than market researchers, is multifaceted, with different people having very different views about what class means.

What about the perceived character of people in different social classes? As Figure 9 reveals, approximately half of all respondents think there is little difference between the classes – except for on the trait of greed, which is widely thought to be a middle-class characteristic.

Among those who take sides, working-class voters are thought to be much friendlier and more honest and, by narrower margins, more patriotic, faithful to their partners and charitable.

Not surprisingly, when Deltapoll asked about characteristics and class, the question produced some big differences based on the self-defined class of the respondents, as Figure 10 shows.

For all six character judgements, people have a better view of their own social class than do people from the opposite one. That said, middle-class people are more likely to see the virtues of friendliness and honesty in the working class than in their own – though by far smaller margins than working-class respondents. Indeed, working-class voters generally have a consistently more favourable view of their own class than middle-class voters do of their own. Perhaps this contrast in perceived virtue is one reason why some people consider themselves working class despite having well-paid white-collar jobs. “Angels” indeed.

How much does all this matter? After the economic and social changes of recent decades, does it still make sense to talk of class divisions in everyday life?

The answer is yes. Almost two-in-three people (64 per cent) think there are “very” or “fairly big” divisions between the lives of middle- and working-class families today. This is down from the 74 per cent who say there were big divisions two or three generations ago. Middle-class respondents detect the greatest decline in class divisions during this period of time (from 72 per cent choosing “big” to describe the division two or three generations ago down to 56 per cent today), but the differences between the two groups is modest.

Britain is still seen as a country divided by class, as shown in Figure 11.

It is worth noting that the views of voters aged over 65, who can draw directly on their own experience of change over the past four decades or more, do not differ much from the electorate as a whole: a big divide of 75 per cent for two or three generations ago that falls to 57 per cent today; and 19 per cent saying a small or no divide for two or three generations ago, a figure that increases to 39 per cent today.

Party support interacts with social class in the way people view these divisions, with middle-class Conservatives seeing the sharpest fall in class divisions although almost half of them say these divisions still persist:

• Self-defined middle-class Conservatives: 70 per cent say class divisions used to be “very” or “fairly big”, a percentage that reduces to 44 per cent for Britain today, down 26 points.

• Self-defined working-class Conservatives: 78 per cent for the past, reducing to 67 per cent today, down 11 points.

• Self-defined middle-class Labour: 79 per cent for the past, reducing to 75 per cent today, down 4 points.

• Self-defined working-class Labour: 81 per cent for the past, reducing to 79 per cent today, down 2 points.

A Different Picture When Class Is Considered Personally

So class remains a big issue for most people … or does it? A further Deltapoll survey explored the way people view the role of social class in their own lives as distinct from the way they look at British society generally. In this case, a very different picture emerges.

Whereas two-in-three voters detect a big division between the lives of middle- and working-class people in Britain today, just one-in-five say it is important “to people like you nowadays” – in other words, to them personally. This contrasts with the 54 per cent who say class mattered “to people like your parents”, as shown in Figure 12.

On both questions, there are only slight differences in the responses of working- and middle-class respondents. However, it is noteworthy that there are marked differences by age. People aged below 25 divide fairly evenly between those who say class is “important” and those who say “unimportant”. Older people feel differently: they divide by three-to-one, saying class is “unimportant” rather than “important”. The longer that time has elapsed since their childhood years, the less class seems to matter to people.

Finally, Deltapoll explored the role that social class plays in the way in which people view their own identity. Figure 13 confirms that social class matters to only a minority of Britons today. Political outlook is equally a minority concern. These data also confirm the central role of people’s upbringings.

It’s worth noting that nationality – whether being British or belonging to one of the constituent countries of the UK – looms far larger for older Britons. Indeed, for people aged over 55, being British is the biggest single factor among the ten that Deltapoll tested.

Although national identity is not the main concern of this report, it is worth noting the way the figures vary by nation:

Percentage who say being British matters a great deal or a fair amount:

• England: 55%

• Wales: 61%

• Scotland: 37%

Percentage who say being English/Welsh/Scottish matters a great deal or a fair amount:

• England: 46%

• Wales: 62%

• Scotland: 58%

The days have long gone when social class lay at the heart of people’s identity, lives and politics. Today it matters to only a minority.

Yet glimpses of its residual effects can still be seen. Most people think that class mattered to their parents’ generation. They anchor their own class identity in their family roots and believe that class still divides Britain today. The echoes of Britain’s past economic and social order have not completely died away.

In our previous report, From Red Walls to Red Bridges: Rebuilding Labour’s Voter Coalition, we showed that occupation has ceased to be a useful guide to the way people vote. In this follow-up report, we have seen how self-defined class still influences the way some people think and vote – most dramatically in the “crossover” groups: people whose own class identity differs from the one derived from their occupational status. However, these factors flow more from voters’ childhood circumstances than their lives today.

All this raises large, long-term questions about Britain’s future. The influence of social class on voters’ choices is likely to fade further as family links to Britain’s industrial past are broken. What then? A pessimist might quote Antonio Gramsci’s warning from almost a century ago: “The old is dying and the new cannot be born”.

To prove the pessimists wrong, the big parties need to heed two specific lessons.

The first is that while social class matters little to the lives of most voters, the Labour and Conservative “brands” are still firmly connected to the divisions of the past, with Labour generally seen as a party for the working class, and the Conservatives for the middle class. Given that many more people say they are working class rather than middle class, this might seem like good news for Labour. The trouble is that having a reputation based on class – any class – risks a party conveying the impression of being a relic from the past, rather than a force for a better future.

The second big lesson for the parties is that few voters these days feel a strong attachment to any of them. People’s politics make as small a contribution as social class to their identity. A survey of party activists would doubtless produce a far higher number. Politicians and local party members campaigning for votes might usefully reflect that for most people, elections are not the climax of years of effort, but interruptions to the normal rhythms and concerns of daily life.

If class-based politics belong to the past rather than the future, and parties need different tunes to march to, what should they be? One option is that parties should become more like consumer brands, competing in the manner of Tesco versus Sainsbury’s or Samsung versus Panasonic. In other words, like rivals seeking to generate a loyal customer base, but ultimately offering a similar service. This notion has been around for some time – for example, in the work of the pioneering social psychologist Hilde Himmelweit in the 1970s and 1980s. Indeed, parties have increasingly employed the techniques of market research and consumer persuasion to design their campaigns.

However, if parties appeal to voters in purely transactional terms, our democracy will be diminished. We shall lose the essential belief that it has a larger purpose and that parties should embrace social benefit and collective progress, not just individual self-interest. Moreover, a healthy democracy requires healthy parties: ones that can inspire people to join them and campaign for them because they stand for principles worth fighting for. How can parties inspire if they compete like rival brands of washing powder?

In its heyday, class-based party politics provided not just an important choice for voters, but big ideas and rallying cries for different views of society. Those ideas have had their day. The main parties need new big ideas if they are not to descend into an ideas-free, purely transactional politics. The intermittent, relatively recent successes of the Liberal Democrats, Green Party and UKIP – and, above all, the SNP in Scotland – warn Labour and the Conservatives of the power of insurgents offering a compelling story of a different future.

To combat this threat, mainstream politics and, especially, progressive politics need to urgently replace old ideological instincts with new answers, and provide a coherent analysis of what must be done – rooted in a new vision for 21st-century Britain.

Some ingredients of a post-class politics are plain to see. They will reach beyond class with an agenda that can appeal to all social groups. They will address the impact of technology, the need to tackle climate change and Britain’s place in Europe. Perennial issues will remain, such as the role of the state in providing public services and fighting poverty, but only a minority of voters see even these as class struggles.

Cross-cutting these challenges are the awkward matters of culture and identity, which too often play into the hands of the nationalist Right. Progressive parties need not only the right policies but also the credible hope of a better life for all that unifies the electorate – instead of what divides it.

One option is to reach back into the era before Karl Marx and reclaim from his baleful influence the original language of radical reform. A novel doctrine was beginning to develop in the 1820s and 1830s. It opposed the power of wealthy landowners that thwarted progress and impoverished millions. Its dream was of a dynamic society that valued managers and entrepreneurs, professionals and labourers. It promoted liberty, fairness and the common interest.

It was called socialism. Perhaps the change in the word’s meaning happened too long ago for it to be reversed. But it is worth noting that the concept of an inclusive, progressive enterprise economy has deep roots. Ideas from the time before Marx could help to shape the character of politics after class.

Peter Kellner

Peter Kellner is a political analyst, visiting scholar at Carnegie Europe and former president of the pioneering online survey-research company YouGov. He was chairman of YouGov from 2001 to 2007, during which time he was named Chairman of the Year in the 2006 Quoted Companies Alliance awards, and president from 2007 to 2016.

He has written for The Times, Sunday Times, Independent, Observer, Evening Standard, New Statesman and Prospect. He has also been a regular contributor to Newsnight, A Week in Politics, Analysis and the BBC’s election-night results programmes. He has written a variety of books and leaflets about politics, elections and public affairs – and was awarded Journalist of the Year in the British Press Awards in 1978.

Peter has also been a visiting fellow at Nuffield College, Oxford, a distinguished visiting fellow at the Policy Studies Institute, London, and served as a member of committees set up by the Economic and Social Research Council to commission research into elections and social exclusion. He has served as a trustee with various charities and also chaired the Royal Commonwealth Society and the National Council for Voluntary Organisations.

Deltapoll questioned 4,292 people online throughout Great Britain between 12 and 17 January (data in Figures 1 to 11), and 4,003 people online between 14 and 17 March (Figures 12 and 13, and data on national identity). The raw data in both surveys were weighted to the profile of all adults.

The full results are available at deltapoll.co.uk