A key element in the story of this election was Labour’s insufficient recognition of the threat it faced in dozens of marginal seats.
Despite now representing fewer constituencies than at any time since 1935, Labour cannot afford to be complacent about the seats it currently holds.
The pattern described in Chapter 1 – of Labour support dropping and remaining at a lower level, followed after some delay by cumulative increases in Conservative vote share – can be seen not only in seats lost in recent elections, but also in those we still hold but which now look much more marginal.
There are 58 seats across the country that would be lost with a further swing of 6 per cent to the Conservatives. There is no evidence, yet to date, that the long term trends described in Part 1 are yet abating.
At the same time, it would be wrong to imagine that Labour’s new areas of concentrated support amount to a new “core” or “base” that it can count on to stick with it.
Many city-based and academic seats were lost by Labour to the Liberal Democrats and even the Conservatives in 2005 and 2010. High levels of switching and tactical voting among these electorates mean Labour cannot take for granted advances it has made in these areas over recent elections.
To win a parliamentary majority at the next election, Labour needs to gain 123 seats on top of its 2019 result. This would require a swing of over 10 per cent. In doing so, the party would need to win seats in every part of the UK and from other parties as well as the Conservatives. For example, without winning seats from the SNP or Plaid Cymru, the swing required in England and Wales would need to be over 12 per cent. If a boundary review goes ahead as now seems certain, it has been estimated that the net effect would probably be to increase the required swings by at least another percentage point.[157]
According to analysis by the Fabian Society[158], on current boundaries, if the seats Labour needs to win are ordered simply by the swing required relative to the 2019 result:
Winning 123 seats would give Labour a majority of one. A majority that could see a radical, reforming Labour government through a full parliamentary term would need to be larger than this, entailing an even more profound electoral advance.
Labour’s advantage over the Conservatives among younger voters and those living in cities might lead us to hope that the Party’s position will be helped by increases in the relative weight of younger cohorts or more urbanised constituencies in future elections.
The effect such trends may take is highly uncertain, however. There is no guarantee that the political preferences and allegiances of today’s younger voters won’t shift as they grow older[161] - indeed, as we show below, there are signs that support for Labour policies among younger voters is highly variable. Nor is there strong evidence that the spread of conurbations to outer suburbs will automatically increase Labour’s vote share in those constituencies.
Even if there are trends that could work in Labour’s favour, they are not moving fast enough to tilt the balance by 2024, and would take several electoral cycles to bring a decisive benefit. One of the most important lessons of this review has been that losing political ground can be cumulative and increasingly difficult to reverse: the loss of councillor networks, closure of MSP or AM offices, and erosion of CLPs’ membership and links with the community, can reinforce each other and reduce the Party’s presence and campaigning capacity below the critical mass needed to maintain visibility and relevance.
This does not mean that the Party should not take a long-term view that looks beyond the immediate electoral cycle. Given the long running trends described in Part One, we should be using data analysis as well as political judgment on the ground, to develop strategy for building our support in constituencies over time horizons that reach beyond the next election cycle.[162]
Chapter 3 cited analysis by Datapraxis of polling data which maps the UK electorate into fourteen different clusters, each with a prevailing set of views, values and behaviours, and different socio-economic, demographic and regional distributions.
Looking at how these groups vote, Datapraxis found that 13 of them, together accounting for 88 per cent of the total electorate, contained significant numbers of Labour voters in 2019.
Four “core” groups together comprised 29 per cent of the total electorate, of whom 50 per cent or more voted Labour in 2019:
These are the Labour voters we are most familiar with, and among whom it should be easiest to rebuild support. However, it is worth stressing that, even in the unlikely scenario of Labour winning the support of all of these voters, it would give it a vote share of less than 30 per cent.
Labour’s core support was augmented in 2019 by three young groups, together comprising 18 per cent of the total electorate, many of whom were eligible to vote for the first time in 2019.
The other key groups whose decisions have determined the outcome of recent elections are six groups of “swing” voters, together making up 41 per cent of the electorate, each of which includes voters who have supported Labour in the past and will be important to building a winning coalition.
Of these, the most decisive in the last election may have been “Older Brexit Swing Voters”, relatively concentrated in the North East, Yorkshire and Humberside, and the Midlands, and less represented in London. In 2017 Labour won the support of 35 per cent of this group - up on 2015 (26 per cent) and 2010 (23 per cent). However, in 2019 support for Labour among this group fell to just 13 per cent.
The final group, labelled “Establishment Tory Brexiteers”, amount to 12 per cent of the electorate, and included no significant numbers of Labour voters in recent election.
Datapraxis found that across all the thirteen groups from whom Labour has been able to draw support, there was little opposition to most of the Party’s key economic policy positions, from redistribution and corporate regulation to nationalisations, a higher minimum wage, and intervention in the housing market and the economy more generally.
This is consistent with data reported in Part One, indicating that many of Labour’s economic policies were broadly supported, individually and in principle. However a critical issue, as we have seen, is the overall credibility and deliverability of the package.
Digging deeper into the responses helps us to see some important patterns and differences within this broad context.
Views of this agenda are particularly variable among the three younger groups: clear majorities of “Young Insta-Progressives” support these positions but most “Young Apathetic Waverers”, and a significant minority of the “Young Disengaged”, say they don’t know or are neutral. This suggests that support for such policies among younger voters is not as solid as might be inferred from headline polling, and reinforces the case for caution in expecting Labour’s job to get easier as these cohorts make up a larger share of the electorate.
On a more optimistic note, attending to the distribution of “strong” support for different policies also offers indications of what may be most useful in winning over significant shares of key “swing” groups:
Geographical patterns will also be worth investigating. Messages around NHS privatisation may be less resonant in Scotland and Wales, for example, given the devolution of relevant policies in those nations.
The cluster analysis conducted by Datapraxis confirms the divergence on social and cultural issues, highlighted in Part One of this report as an underlying tension and weakness in Labour’s potential coalition.
It is important to recognise that there are areas of relative consensus - some of them hard fought for. Equal marriage is comparatively uncontroversial, consistent with other polling data indicating that views around sexuality, as well as gender and race, have converged over recent decades.
A key fault line runs between the three most “socially liberal” groups - “Progressive Cosmopolitans”, “Younger Instagram Progressives”, and “The Green Left” - and the rest of the electorate on many currently salient issues:
There is a range of views within every group, so these divisions are not absolute. However, the broad analysis makes clear the challenge Labour faces in this area: the three groups with the most “socially liberal” views across all questions make up the majority of Labour’s “core vote”, but just 21 per cent of the total electorate - barely half of the proportion needed to achieve a winning vote share. Disagreement with these three groups on some social issues is particularly sharp among the three “Brexit swing groups”, who are closest to them on many economic issues and are likely to be key to winning back seats in the North and Midlands.
Managing these differences will require a careful mediation, understanding and bridge-building between different groups, while not departing from Labour’s values and principles. As John Denham, who held a marginal seat for Labour for many years, has observed in the past “Labour does not need to agree with the voters on everything, just that we should not disagree with them on everything”. How this is achieved in terms of Labour policy is not a matter for this review; however, in the next chapter we set out how the Commission believes the party should strategically approach the challenge of coalition building between these different groups.