We won't know in more detail until some polls on "why did you vote that way" are published (and I haven't found any insofar, but some were published after the 2017 election, so no doubt some will be published for this one too), however what we do know are some correlations with the Brexit vote:
Labour lost votes in both strong Remain and strong Leave areas.
Strong Leave and strong Remain constituencies are those where an estimated 60% or more of the electorate voted for that option at the EU referendum.
These estimates of constituency Brexit votes were modelled by Professor Chris Hanretty, as the 2016 referendum result was only recorded by local authority and not by Westminster constituency.
The Conservatives were clear winners in constituencies estimated to have voted majority Leave in 2016. They won almost three quarters of all these seats.
By contrast, there was no clear winner among Remain backing constituencies, with a crowded field of parties all winning substantial numbers of seats. [...]
Labour did best of all those parties but only took 40% of the constituencies that backed Remain.
So yes, Brexit apparently had a role in the Labour losses this year and it may have been a "double whammy", with losses against Conservatives in the Leave areas and inability to win against other [clearer] Remain parties in the Remain areas...
The other noticeable thing, although probably correlated with Brexit Leave vote (so not necessarily distinct from it), is the unconvincing performance of Labour in working class areas...
Overall, the Conservatives broke new ground, moving into many traditional Labour heartlands.
In 2017, Labour held 72 of the 100 constituencies with the most working class households (defined as C2DE using data from the 2011 census).
In 2019, this figure fell to 53 and the Conservatives increased their share from 13 to 31.
An FT analysis (as quoted by Vox) claims this is the strongest association with the voting pattern:
“In seats with high shares of people in low-skilled jobs, the Conservative vote share increased by an average of six percentage points and the Labour share fell by 14 points. In seats with the lowest share of low-skilled jobs, the Tory vote share fell by four points and Labour’s fell by seven,” the FT said in its analysis. “The swing of working class areas from Labour to Conservative had the strongest statistical association of any explored by the FT.”
However, the same article notes:
This is extremely preliminary: We don’t yet know which voters in these constituencies voted which way, so we can’t yet say whether class itself is the key. Indeed, another analysis by Will Jennings, a political scientist at the University of Southampton, suggested that education level — the percentage of college graduates in a constituency — was actually more important than income level or class per se, which would be consistent with long-term data on European political realignment.
But it’s also clear that Labour did badly across the board: Jennings’s analysis finds that Labour lost support even in cities, a result that suggests that Corbyn’s personal unpopularity was depressing voters who should (on the Brexit theory) be supporting the more Remain-friendly party.
Also of note is that in Scotland both Labour and the Conservatives lost substantially to the SNP, although Labour got nearly wiped out (from 7 to 1 seat; Conservatives from 13 to 6). Whether the SNP surge is due to pure Scottish nationalism or in combination with Brexit concerns is also not terribly clear right now, without additional polling into voters' motivation. However, as far as Scotland goes, the poor performance of Labour (as well as the SNP surge) was nearly a repeat of the MEP elections held earlier this year. Interestingly, the SNP however is below its 2015 heyday in Scotland, this year, due to Conservatives and Liberal Democrats doing better compared to 2015.
YouGov in contrast found that age was the strongest demographic predictor of the vote this year as well, and that the "cutoff" has drifted by some 8 years, compared to the last election:
In the largest survey of the election so far, YouGov found age is still the biggest dividing line in British politics today. [...]
The Labour to Conservative crossover is now at 39, which is down from 47 at the last election in 2017.
YouGov also found a substantial gender gap among the young, with young women even more likely to vote for Labour, compared to young men.
Interestingly, it seems Labour lost votes among the highly educated in favor of Lib Dems and in favor of Conservatives among the less educated:
The highest level of education someone has achieved remains an important dividing line in how people vote. Labour did much better than the Conservatives amongst those who have a degree or higher, by 43% to 29%.
The Liberal Democrats also performed very well amongst this group with 17% of the vote share. We saw in 2016 that those with a higher education level were overwhelmingly more likely to back remaining in the EU, and this has seemingly transferred into party voting.
Compared to 2017, the Conservatives have improved amongst those without a degree, but performed worse amongst those with a degree or higher. Labour lost voters amongst all three education level groups.
Overall YouGov found that Labour lost 2017 voters to both Conservatives and to Lib Dems, but that those lost to the latter failed to transform into seat gains for the LibDems (even though the Lib Dems did better in terms of the popular vote than in 2017)
YouGov has a similar picture in terms of the 2016 referendum prior:
This time around the Conservatives managed to boost their vote share amongst Leave voters to three quarters (74%) while the Labour Party actually reduced their share of Remain voters to just under half (49%). This fall came mainly at the hands of the Liberal Democrats, who increased their vote share amongst Remain voters to 21%, compared to 12% in 2017.
An older YouGov poll on Nov 5 probably explains well enough the electorate's confusion over Labour's Brexit stance (this might have been alleviated somewhat since then, but probably not dramatically):
With 70 per cent of Brits saying Brexit is the biggest issue facing the UK, a whopping two-thirds of British voters said they are “unclear” about Labour’s Brexit position.
Just 21 per cent told Yougov that Labour’s stance made sense to them, with Corbyn’s non-committal stance leading to a lack of clarity among voters.
Furthermore, 65 per cent of Remainers could not work out Labour’s thinking on Brexit, and 57 per cent of people who backed Labour in 2017’s snap general election found the party’s latest stance unclear. [...]
In comparison, 57 per cent of voters said Boris Johnson’s Tory party was clear about its Brexit policy, compared to 29 per cent who are confused about it.
And in an early December poll focused on the same issue:
New YouGov data shows that just one in five (21%) Labour Leavers think a second referendum where the Government remains neutral would be a good outcome, rising to 36% when including those who say it would be an acceptable compromise. [...]
Further bad news for Corbyn is that his second referendum proposal isn’t even the most favoured Brexit option with Labour Remainers. Some 63% think this is a good outcome, slightly fewer than the 70% who say revoking Article 50 would be a good outcome.
So I think it's fair to say that whatever other issue weighed on the electorate's mind, the Labour's final approach to Brexit was at best a "second best" choice to many in both the Leave and Remain camps...
YouGov did run a pre-election poll, but quite early on, in October, on which issues the voters were considering the most important:
Besides some perhaps interesting changes they chose to highlight, Brexit was clearly the dominant issue in many voters' minds... but then so it was in 2017.
Oddly, an Ipsos Mori poll put quite a different number on "health" since they asked the question in terms of NHS; it might also be because of the debate and rumours later in the campaign, as this Ipsos poll ran approximately one month later.
It would be interesting to know if the NHS emphasis backfired on Labour, but it's impossible to tell from these polls.
There's one Dec 9 poll by YouGov that found that both parties' manifesto were remembered by the electorate for their NHS promises equally (22%), although in the case of Labour NHS was the most notable issue in their manifesto, whereas for the Conservatives their Brexit promise held that spot (43%). In terms of manifesto credibility, that poll only asked an omnibus question and found that Conservatives had the edge:
However, only 20% of Britons think that Labour’s policies are well thought through, compared to 27% who say the same of Conservative policies.
Lord Ashcroft ran an exit poll which also asked about issues. Interestingly, the NHS topped the list overall, but not for Conservative voters for whom "getting Brexit done" was the top issue. (Actually, if we add the pro and contra-Brexit issues, which are counted separately by Ashcroft's poll, they do edge the NHS overall.)
Euronews has a somewhat different explanation, quoted from a Ipsos Mori expert; basically, the Remain vote was more split in this election than in the previous one:
in terms of the popular vote, the Tories' share increased by a mere 1.2 per cent compared to the last election. Compared to his predecessor Theresa May, only 270,000 more people voted for Boris Johnson's Conservative Party on Thursday, official results showed.
Keiran Pedley, Research Director of Public Affairs at IPSOS Mori, told Euronews that the key factor behind the Conservatives' sweeping victory was actually the losses recorded by the Labour Party.
"Although the Conservative vote stayed reasonably steady, nationally Labour's vote share dropped by 8 percentage points," Pedley noted. [...]
Pedley said that it was fair to say a split remain vote had contributed to hand Boris Johnson a victory.
"It was always quite likely to be the case because the nature of the remain vote is that it's spread across different parties -- so at a very basic level, Labour, the Liberal Democrats, the Green Party, the SNP, etc." the expert told Euronews.
"There is analysis out today suggesting that if you just added up parties supporting a second referendum on Brexit versus those who oppose, ironically the parties that support a second referendum would get 52 per cent of the vote and the parties that oppose would get 48 per cent of the vote."
But the expert warned that the split Remain vote was not the only explanation. [...]
Whereas social class used to be a key dividing line in how people vote, the expert said other factors had now taken over - including age, education, urban vs rural areas, etc.
But they don't offer much detail on the latter.
The New York Times offers an interesting parallel though with Trump's election:
By Friday morning, Britons awoke to a Labour Party largely consigned to the cities of England. The Conservatives, on the other hand, harnessed the power of Brexit to storm districts where the party’s brand had been toxic for generations.
In doing so, they replicated the success of President Trump in breaching the so-called Blue Wall in states like Michigan and Wisconsin in 2016, exploiting a combination of anti-immigrant messaging and dissolving class allegiances to take seats thought to belong to the Democrats.
The NYT article has some interviews with voters in these non-city areas that switched to the Conservatives mainly because of Brexit, but it's obviously rather anecdotal evidence. The NYT article also ponders whether this change/realignment is permanent. Its author seems unaware of some BES studies that show increased voter volatility in the UK in the past decade, e.g "Across the three elections,
2010/15/17, 49% of people, almost half, did not vote for the same party each time."
The Spectator has publicized a poll by Opinium, which found that the main reason for people switching vote away from Labour was its leadership, i.e. Corbyn, although the party's stance on Brexit follows as the 2nd most mentioned reason:
Of those who backed the party in 2017 but failed to vote Labour this time around, 37 per cent of them said it was because of the leadership – compare that to the 21 per cent who blamed the party’s stance on Brexit. The results of that 5,641 person poll are set out below in all their grisly detail:
The same Opinium poll was actually published by a number of right-leaning media, including The Sun and Daily Mail, but seemingly not mentioned in the more left-leaning outlets.
A pre-election poll by Deltapoll and mentioned by The Independent found similarly:
Another survey by Delta Poll which asked potential Labour defectors why they might vote for someone else or – more likely in an election where the party failed to turn out much of its base – for no one at all. Fully 46 per cent cited Jeremy Corbyn as a reason, compared to just 19 per cent who said it was about Brexit.
This latter poll was "Prepared by Deltapoll for the Jewish