Labour’s electoral demise has been long predicted. ‘If necessary changes are not made, the Labour vote will probably decline … by about two per cent at each successive general election’ wrote Tony Crosland in 1960. In the same year, two leading academics, Richard Rose and Mark Abrams, published the much discussed book: ‘Must Labour lose?’ Within four years Harold Wilson replied ‘no’ by winning, just, in 1964. A quarter of a century later Tony Blair did the same, much more decisively.
So is the party’s 2015 defeat just another in the repeating cycle of rise and fall? Or has something more fundamental happened this time that means Labour’s plight is both desperate and lasting? Unfortunately close psephological study of the 2015 result suggests the latter. In truth had Labour chosen in the recent contest a leader who combined the best qualities of Keir Hardie, Clem Attlee and Jim Callaghan – I am not entirely sure that one was available – it would still have been in crisis.
How bad is Labour’s plight? Of course, the result fell short of expectations. These were fuelled by statistically unsound opinion polls. The basic truth is that even if you have a perfect sample 19 in 20 polls have a margin of error of 2-3 points on each party’s share, with one in 20 having a still bigger margin. So ‘Labour 33, Tory 33’ could statistically be anything from ‘Labour 36, Tory 30’ or ‘Labour 30, Tory 36’ which puts a rather different complexion on polls as a guide. As Patrick Sturgis’ inquiry into the polls is showing, the odd thing is that the polls all clustered around the same results – but that just mean they were all wrong. The truth probably is that this was an election that Labour was never going to win.
Still, the Tories have an overall majority of only 11. In England, there was actually a small swing to Labour. One more crisis of capitalism, such as occurred after 2008 and Labour might surely be swept into power, so the optimists’ argument goes
Only, as the psephologists rake over the embers of May’s fire, spontaneous recovery looks less and less likely. In September Ron Johnston of Bristol university addressed the all party statistics group, his work with David Rossiter and Charles Pattie unveils the unplumbed depths of Labour’s plight.
First, the party lost 40 seats in Scotland to the Scottish National party. Of course the SNP could falter, but its underlying political position is a strong one. Following the referendum, for Scots arguably the optimum outcome is to remain in the union, while using the threat of withdrawal to extract endless subsidy and limitless devolution out of Westminster. Canny Scots will vote accordingly.
The bouleversement in Scotland is part of the reason why the bias in the electoral system has turned against Labour. Bias is a measure of who does best under the electoral system, by calculating how many seats each big party would have if they had an equal percentage share of the national vote. From Blair’s time onward, that bias has been strongly pro-Labour – over 100 seats in both 2001 and 2005, though rather reduced since. Johnston calculates it as having been worth 64 seats to Labour in 2010 and he expected before the 2015 election that it would be worth 35 seats in 2015. In fact the 2015 bias was not 35 to Labour but 59 to the Tories.
This was partly due to Scotland where large numbers of ‘wasted’ votes for Labour yielded just one seat. But not entirely so, for in England there was a similar phenomenon. Labour piled up big majorities in its safe seats without adding to its total tally of seats. This was in large part because of the collapse of the Liberal Democrat vote in the seats they formerly contested with Labour but where in 2015 they came either a poor second or even worse. However, it did poorly in the marginal constituencies held by the Conservatives, with the result that, in England and Wales alone, the bias in favour of the Tories amounted to 28 seats.
The result is summed up in The British General Election of 2015, the authoritative academic study by Philip Cowley and Dennis Kavanagh. It shows that, if the two main parties had tie in votes in 2015, the Tories would have had 302 seats to Labour’s 256. To win more seats than the Tories at the general election, Labour would have had to poll 3.7 per cent more votes than the Tories. A Labour overall Commons majority would have required a 12.5 per cent Labour lead in votes.
The mountain Labour has to climb will get yet steeper. In the last parliament the Liberal Democrats scuppered the Tory attempt to push through a constituency redistribution that would have meant a net swing of seats to the Tories of around 20. That redistribution with probably the same consequences is highly likely to go through for 2020. In addition the early introduction of individual voter registration is likely to reduce the number of people able to vote, likely excluding disproportionately young urban would be Labour voters.
The number of older people in the population continues to rise so that according to the Fabian Society by 2020 more than half of the electorate will be 55 or older. Already in 2015 ‘It has been estimated that the Tories’ lead in votes from the over-65s is equivalent to their total majority’, according to Labour’s official report on its defeat chaired by Margaret Beckett. Older people are more likely to be on the register as they move home less, more likely to vote and, sadly, more likely to vote Tory. So unless Labour can overcome its age deficit, it will struggle to gain ground.
Finally, fatally, the number of marginal seats – the seats Labour needs to win – has shrunk. With the number of seats being reduced from 650 to 600 the exact number of gains Labour needs is uncertain. But if the next election were fought on the 2015 boundaries – ignoring the adverse effects of the boundary changes – Labour would need to gain 93 seats to command an overall majority. That would require a 9.5 per cent swing from 2015.
There simply are not the marginal available now to make that plausible. Johnston argues that in 2010 the number of Tory seats with a 10 per cent or less Tory lead, and thus vulnerable to a 5=five per cent swing, was 81. After 2015 there are only 56 such seats.
Equally 50 further seats are vulnerable to a swing of five per cent to the Tories from 2015. If they were all lost the party would be pushed back to holding only 182 seats, 27 fewer than it won at the disastrous 1983 general election. The picture that would then have emerged is of the Labour party penned into its remaining traditional strongholds while losing out in the seats that really matter for forming a government.
That was the situation before the election of Jeremy Corbyn. Theoretically of course a public wave of enthusiasm for the new leader following the wave of enthusiasm among the new Corbynistas in the party could transform the electoral situation. It has to be said that there is precious little sign of that so far.
At a time when the government should be entering the usual period of unpopularity the polls, for what they are worth, have swung slightly against Labour. And as the objective Cowley and Kavanagh say in the election study: ‘Almost all psephological analysis of Labour’s support between 2010 and 2015, as well as what we know about non-voters, the Ukip vote or indeed the nature of support for the Scottish National party, would indicate that the Corbyn strategy is a route to an electoral brick wall’.
David Lipsey is a fellow of the Academy of Social Sciences and joint chair of the All Party Statistics Group.
Progressive centre-ground Labour politics does not come for free.
Our work depends on you.