This is the first in a series of posts on Jeremy Corbyn’s leadership campaign and the issues it raises.
The charge most commonly thrown at Corbyn from the Right is that Labour couldn’t possibly win in 2020 under his leadership. So I’m going to start by looking at how Labour’s won before, and at trends in voting at General Elections more generally. I’m going to argue that vote shifts in General Elections since 1945 can be modelled using a reasonably small set of effects:
1. Straight Vote Switch
2. Mobilisation Effects
2.1. Selective Mobilisation (Benefiting Incumbent)
2.2. Selective Mobilisation (Benefiting Opposition)
2.3. Selective Demobilisation (Benefiting Incumbent)
2.4. Selective Demobilisation (Benefiting Opposition)
2.5. General Demobilisation
(I haven’t got any examples of general mobilisation.)
3. Minor-Party Effects
3.1. Minor Party Surge (Benefiting Incumbent)
3.2. Minor Party Surge (Benefiting Opposition)
3.3. Minor Party Squeeze (Benefiting Incumbent)
3.4. Minor Party Squeeze (Benefiting Opposition)
4. Incumbency Effects
4.1. Consolidation In Office
4.2. Fightback In Opposition
Disclaimers: I’m not a political scientist or a historian of the period (although I have lived through most of it); for the most part I’m going to be ignoring day-to-day or even month-to-month politics and discounting the two- or three-way shifts in loyalty which take place in reality. It’s a simplistic model, but I think we can make it talk.
Labour won the 1945 election with a majority of 146, gained from 12 million votes – just under 48% of the total votes cast, and more than the votes gained by the Tories and Liberals combined. (The thin green line you can just make out in the left-most column, incidentally, represents 300,000 votes cast for the Communist Party and the left-wing Common Wealth party – votes which elected two MPs and one, respectively.) The figure to focus on here is the 8.9 million non-voters (turnout was 72.8%). In 1950 the voting public made up a significantly larger proportion (83.9%) of a larger electorate. Labour picked up some of the increase, but the Tories picked up more; Labour won with an overall majority of five. This is the first pattern I want to highlight: (2.2) Selective Mobilisation. Mobilisation in this sense isn’t a matter of winning voters from one party to another, or even ‘getting the vote out’ in the door-knocking and lift-to-the-polls sense, but of motivating potential voters: making the political weather to the point where voting for a particular party seems sensible. The point is simply to persuade supporters to vote rather than not bothering; getting them to feel that turning out to vote is a good idea, even if it hadn’t seemed to be before. Given the increased size of the electorate it’s hard to be certain where the votes for any party came from, but the fact that abstentions fell by 3.4 million and the Tory vote rose by 2.5 million looks decidedly suggestive.
The second pattern I want to draw attention to makes its appearance in 1951, when the Attlee government ill-advisedly called an election in the hope of increasing its majority. Labour certainly increased its vote, winning the most votes the party had (and has) ever won, but the Tories increased theirs more and did so more effectively. The key mechanism here was the (3.4) Minor Party Squeeze. The 1.8 million votes shown here for Liberals includes 1.1 million votes for the ‘National Liberals’, a Tory-allied splinter dating back to Ramsay MacDonald’s coalition government. The National Liberal vote went up slightly between 1950 and 1951, but the true Liberal vote plummeted from 2.6 million to 700,000. A third-party squeeze allowed the second party to achieve a gain of 22 seats – and a change of government – even while the Labour vote increased; Labour vote share in 1951 was 48.8%, even higher than it had been in 1945.
If we look at the next two elections – 1955 and 1959 – two patterns are discernible. One is (4.1) Consolidation In Office: over three successive elections the Tory vote grows, little by little, and the Labour vote declines. The other, particularly apparent in 1955, is (2.3) Selective Demobilisation, benefiting the party in office. The reverse of mobilisation, this – again – isn’t a matter of persuading opposition voters to switch parties, but simply demoralising them to the point where they stay at home. (I’m not saying that this was Tory party strategy – or even that anyone set out to achieve it; the main agents of Labour voter demoralisation may well have been the Labour Party. I’m just saying that the figures seem to suggest that it happened: electorate up by half a million or less, non-voters up by 2.1 million, Labour vote down by 1.5 million.) Patterns 3 and 4 are both essentially Labour/Tory phenomena in these years; the Liberal vote is unaffected, holding up between 1951 and 1955, then growing markedly in 1959 under the incoming leadership of Jo Grimond. (What looks like a collapse in 1955 is down to the almost total dissolution of the National Liberals into the Tory Party.)
The Tories’ reign came to an end in 1964, when the Labour vote leapt from 12.2 million to, er, 12.2 million; the national vote was actually 10,000 lower in 1964 than 1959. What had changed, however, was the Tory/Liberal share of the vote – or, to be more precise, the Tory/Liberal/non-voter share of voters. Here we see another pattern: the (3.2) Minor Party Surge, cracking the political pack-ice to the benefit of the main opposition party (in this case, Labour). Which in turn called an early election in 1966, pulling off a classic example of (4.1) Consolidation In Office by taking votes both from the Liberals and directly from the Tories.
What happened next?
The size of the electorate jumped between 1966 and 1970, due to Wilson’s government giving 18-year-olds the vote. As you can see, Labour didn’t gain from this. In fact none of the patterns identified up to now really fits the way that Edward Heath’s Tory government came to power. Let’s just call it a (1) Straight Vote Switch – the simplest and (one might assume) most common way for an election to change things in a two-party system, appearing now for the first time in seven elections.
After the February 1974 election, Harold Wilson formed a minority government reliant on, among others, the Ulster Unionists, who had just broken with the Tories. What made it possible was, once again, a (3.2) Minor Party Surge to the benefit of the opposition. (Thanks again, Liberals!) The October 1974 election was called in an attempt at (4.1) Consolidation In Office; unfortunately the Labour vote actually dropped. However, the Tory vote dropped by a lot more; Labour achieved the desired result – a parliamentary majority – through (2.3) Selective Demobilisation. (Labour’s majority was 3 – smaller than the majority which had prompted Attlee to call the 1951 election.) This was also the period of the SNP’s first surge, from 100,000 votes in 1966 to 300,000 in 1970, 600,000 in February 1974 and now 800,000. The devolution referendum in 1979, closely followed by Margaret Thatcher’s election, would put this into reverse.
As for Thatcher, what these figures suggest very strongly is that her victory in 1979 was almost entirely a question of (2.2) Selective Mobilisation, with a bit of (3.4) Minor Party Squeeze on the side. The Tories may have taken votes from the Liberals and the SNP, but there is no sign that they took any directly from Labour; the Labour vote actually went up compared to October 1974. (An alternative reading is that a Labour->Tory outflow was balanced by SNP->Labour and Liberal->Labour inflows. Either way, Labour didn’t get any less popular.) The main contributor to the massive increase in the Tory vote – from 10.5 million to 13.7 million – seems to have been a drop of nearly 2 million in the number of non-voters. The thin dark blue line you can make out near the top of the column, incidentally, is Britain’s first substantial far-Right vote: 200,000 votes for the National Front.
Bring on phase three:
Research published recently has argued that the ‘Falklands Factor’ had very little effect on the Tories’ poll ratings, and had dissipated by the time of the election the following year. If anything accounted for Thatcher’s second victory, the paper argued, it was our old friend the economy, and people’s subjective perceptions that their prospects were improving. I’m happy to bid the Falklands farewell, but I’m not sure that the voting figures support the second argument. What leaps out is the huge success of the SDP/Liberal Alliance, who put on 3.5 million votes relative to the Liberals’ vote in 1979; Labour’s vote, meanwhile, fell by 3 million compared to four years earlier. The Tories’ vote actually fell, as did overall turnout (from 76% back down to the 73% of October 1974). In short, what we’re looking at here is a rare example of (3.1) Minor Party Surge to the benefit of the incumbent.
The next couple of elections are interesting (to look back on; they were heartbreaking to live through). There’s (4.1) Consolidation In Office, with the Tory vote increasing at both elections; John Major’s 14.1 million is the highest vote ever obtained by any UK political party. But there’s also (4.2) Fightback In Opposition, with the Labour vote also increasing both times – and by substantially more. (The Tories’ share of the vote drops slightly over the period, from 42.4% to 41.9%. Labour’s share increases from 27.6% to 34.4%.) These increases are paid for by a combination of (2.1/2) Selective Mobilisation (benefiting both the leading parties – although Labour more than the Tories) and (3.4) Minor Party Squeeze (in Labour’s favour). Politics over the period was getting steadily more interesting – only 9.5 million non-voters in 1992, the lowest figure since February 1974 – and more polarised between the two main parties; and the balance between the two was steadily shifting towards Labour.
Then came 1997, which seemed to represent a triumphal culmination to the growth of Labour in opposition. Actually, as we can see, it reversed most of the trends which had been operating. The (3.4) Minor Party Squeeze continued to work for Labour, but it was accompanied by something that hadn’t been seen since 1970 – a large-scale (1) Straight Vote Switch – and another shift which hadn’t been seen at all since the War: (2.4) Selective Demobilisation of the incumbent party’s support. The increase in Labour’s vote (1.9 million) was huge; the drop in Tory support was more than twice as big. For those of us who remember this period, this both is and isn’t surprising. We can certainly remember the wheels coming off the Tory Party: John Major’s genius move, putting Thatcher’s transformation of the political landscape in the bank and fronting it with the appearance of reasoned moderation, stopped working more or less as soon as he had to start coming up with policies of his own. The sense that absolutely tons of people were voting Labour nowadays – and that hardly anyone was voting Tory any more – was certainly in the air that May. But look at that huge increase in non-voters – and look at the size of that dark blue stripe (100,000 UKIP votes and 800,000 for Jimmy Goldsmith’s Referendum Party). The idea that Tony Blair’s success, right from the outset, might have depended in part on encouraging a lot of Labour’s opponents not to bother voting – and, perhaps, encouraging another slice of people to go right to the extreme and cast a harmless ‘expressive’ vote – is unpleasantly thought-provoking. All the more so in the light of what happened in phase four:
Just look at 2001. Ladies and gentlemen, I give you, for the first time in a British post-war General Election, (2.5) General Demobilisation. The Greens’ vote had more than doubled and UKIP’s nearly quadrupled – although this didn’t make up for the loss of the Referendum Party – but all the major parties showed big declines in their vote as compared to 1997: the Lib Dems were down 10%, the Tories 12% and Labour 20%. Not only was Labour’s vote was below the level at which Kinnock had lost in 1992; it was below the number of votes gained by Callaghan’s Labour Party in 1979, despite the electorate having grown by nearly 10% over those 22 years. The number of non-voters was unprecedentedly high – for the first time ever, non-voters outnumbered voters for the winning party. Helped, perhaps, by the former more than the latter, Labour secured a second three-figure majority.
At the 2005 election things got still worse for Labour: general demobilisation was replaced by (2.2) Selective Mobilisation of both the other two parties, together with a (1) Straight Vote Switch away from Labour, for reasons that don’t need repeating here. The Lib Dems, who opposed the Iraq war, were the main beneficiary; they gained 1.2 million votes where Labour lost 1.1 million. (The number of non-voters also fell by a million in this period, however; it’s impossible to identify flows with any certainty from data at this level of generality.) There were also the early signs of an impending (3.2) Minor Party Surge from the Kippers. In 2010, epoch-making election though it was, nothing much actually changed. That is, exactly the same things happened as had happened in 2005: the number of non-voters fell; the Tory, Lib Dem and UKIP votes rose; and the Labour vote (already in 2005 the second lowest since the War) fell again. The only significant change was that, in this period, the Tories were more effective than the Lib Dems in re-mobilising their dormant vote (and, perhaps, attracting Labour voters); the two parties’ votes grew by 20% and 13% respectively between 2005 and 2010, as opposed to 5% and 25% between 2001 and 2005. Labour’s vote fell by 10% in both periods. If 2010 was the end of New Labour, then – judging purely in terms of electoral success – New Labour left the party in an appallingly bad state.
The 2015 election was the first since 1997 when the Labour Party’s vote actually increased relative to the previous General Election. So that’s a (4.2) Fightback In Opposition. Unfortunately for Labour it was accompanied by (4.1) Consolidation In Office, along with the bizarre and unprecedented combination of a (3.3) Minor Party Squeeze (Benefiting Incumbent) with a (3.1) Minor Party Surge (Benefiting Incumbent). To be precise, there were three separate minor party surges; while UKIP will certainly have cost the Tories votes, the other two hurt Labour far more. The non-Labour centre-Left wind is blowing far harder against Labour now than it ever did against New Labour (for reasons that may not be mysterious); if we assume that half of the increase in Green, SNP and Plaid Cymru votes in 2015 came from erstwhile Labour voters (a fairly conservative estimate), then Labour effectively needed to gain a million votes just to stand still. Labour actually put on 700,000 votes. For comparison, in 1997 – the only New Labour election which is comparable, as Labour lost votes in all the others – Labour’s vote was up by 1.9 million on the previous election; the Greens and the two nationalist parties between them were down by 100,000 votes. The dog that didn’t bark this year, at least as far as the overall figures can tell us, was mobilisation; a lot of people have got out of the habit of voting since 1997, and the 2015 vote wasn’t enough to get them back into it. Here again there may be a more complex picture if we drill down – mobilisation in Scotland, demobilisation in Lib Dem country? – but the overall picture is static.
So, what have we learnt?
Why Do Governments Fall?
That’s a very interesting question, which I’ll answer if I may by pointing out that nobody has any idea. But my simplistic model does suggest that, on the seven occasions when the government has changed hands since 1945, the following factors have been at work in vote changes:
1. Straight Vote Switch (1970, 1997 and 2010): voters for one party switch to another
2.2. Selective Mobilisation (1979 and 2010): ‘dormant’ voters for the main opposition party become more likely to vote
2.4. Selective Demobilisation (1997): voters for the incumbent become less likely to vote
3.2. Minor Party Surge (1964, 1974, 2010): a rise in support for a minor party cuts away the incumbent’s base
3.4. Minor Party Squeeze (1951, 1979, 1997): the main opposition party poaches support from a minor party
However, most of these have also been a factor on one on the ten occasions when government hasn’t changed hands. The only ones which haven’t are
2.4. Selective Demobilisation
3.2. Minor Party Surge
Which seems to suggest that the best way to win an election is to join a different party. Politics is hard.
We can draw a few conclusions, though. Here are four.
1. Vote Switches Are Rare
In 17 elections (from 1950 to 2015), large-scale vote switching from party A to party B is only identifiable on four occasions – 1970, 1997, 2005 and 2010. One of those didn’t lead to a change of government; out of the other three, 1970 is the only case where large-scale vote-switching is the only identifiable factor.
2. Minor Parties Are Crucial
A minor party squeeze is identifiable in six of these elections – 1951, 1979, 1987, 1992, 1997 and 2015; a minor party surge in five (1964, 1974, 1983, 2010 and 2015 again). Six of these ten elections led to a change of government.
3. Mobilisation Is Important
Selective mobilisation – which rests, not on getting the vote out, but on ‘making the weather’ in your party’s favour; doing the Gramscian thing and getting your ideas into the common sense of the age – seems to have been a factor in 1950, 1979, 1987, 1992, 2005 and 2010. Admittedly only two of those elections led to a change of government, but all the other four either laid the groundwork for a change of government or seemed to at the time (1997, as I’ve argued above, was as much a break with the earlier trend as a continuation). The story of demobilisation is more interesting. It’s at work in 1955, 1959 and October 1974, in each case helping consolidate a previous election victory. It’s turned against the incumbent in 1997, and effectively goes viral in 2001; the total number of non-voters never reached 12 million before 1997, and has never fallen below 15 million since 2001.
4. New Labour Was Weird
For Labour to win a landslide victory, on a wave of public euphoria, with promises to transform the entire political landscape, on the back of a substantial fall in turnout (from 77.7% to 71.6%) was, in retrospect, odd. Following this with a second landslide victory on the back of an even bigger decline in turnout was very odd indeed, particularly when Labour’s vote fell considerably more than any of the other parties’. And gaining a third victory – not a landslide this time, admittedly, but a very substantial majority – on an even smaller number of votes, when both overall turnout and the other major parties’ votes were starting to pick up; that was downright flukey. Neither Blair nor anyone else was going to ride that kind of luck to a fourth election victory. Perhaps 2010 was New Labour’s Best Before date.
Still, New Labour did make some enduring changes to British politics, and I fear that lower turnout may turn out to be one of them. In an odd way there may be some truth in the absurd story put about by some Labour people after the 1983 election – that people hadn’t bothered to vote because they were so happy with how things were going. Part of the positive message of New Labour was that there was a whole new approach to doing politics – an approach which didn’t have anything to do with class conflict (or any other kind of conflict), which promoted a combination of practitioner expertise and scientific management techniques, and which generally looked a lot like managerialism. It would be easy to take from that the message that politics wasn’t something ordinary people needed to worry about – the machine would go on working, in much the same way, whether we tried to get involved or not. The negative message of New Labour, on the other hand, was that this new way of doing politics was going to be the only game in town whether you liked it or not; if you weren’t going to be part of the solution, well, sod you. I think a lot of people – mostly but by no means exclusively on the Right – picked up on this and thought well, sod you then. And gave up on voting – either for good, or just until a “sod the lot of ’em” candidate came along.
Maybe managerialism on one side and disengagement on the other is the modern (post-modern?) condition; maybe weird is the new normal. Or maybe New Labour is over; maybe the belief that, underneath it all, the elitist managerial approach to politics was about something has gone for good. Maybe the only way to win elections on that basis is to be cynical, divisive, dishonest and lucky. In which case Labour is definitely going to need some new tricks for 2020.
Why The Post-War History Of Britain Shows That We Must Support My Politics
Can Labour win again? It’s going to be hard to win any kind of majority on the basis of simple vote-switching; the Labour vote is just too low – and the Tory vote is still well below its Thatcher-era highs, suggesting that 2015 Tory voters are likely to be relatively hard to detach. If we imagine the Labour vote going up by two million, entirely at the expense of the Tories – which is more than Labour achieved in 1997 – the resultant vote would put Labour on the level of 1992, a whisker ahead of 1979.
Labour needs to think much more strategically. Tory-to-Labour switches are nice to have, but what the party really needs is the reversal of the other main trends at work in 2010 and 2015. In other words, what’s needed is a (3.4) Minor Party Squeeze, a (3.2) Minor Party Surge and (2.4) Selective Demobilisation. Firstly, Labour needs to win (back) votes from the SNP and (to a lesser extent) the Greens; if we can win votes back from UKIP as well, so much the better. Secondly, we need a strong Liberal Democrat Party – but one that’s strong against the Tories, as it was from 1997 to 2010. Thirdly, while I’m loath to discourage anyone from voting, it would help Labour a lot if people leaning towards the Tories were that much less likely to vote; if, when you said ‘Vote’, people tended to complete the sentence with ‘Labour’. At the last election the precise opposite seems to have been the case, with the effects that we know (on the polls as well as the result).
The question then is, what kind of party is going to be able to do those things and/or foster the conditions in which those things happen? Whose approach will be better at winning voters back from the Greens and the SNP – a Labourite hack who sees one lot as tree-huggers and the other as tartan Tories, or a principled socialist who sees them both as friendly rivals? Which approach will do less damage to Tim Farron’s crusade to retake the West Country – scrapping over the middle ground and denouncing the Lib Dems as soft on drugs and civil liberties, or seeing them as nice well-meaning halfway-house merchants and leaving them to it? What’s the best way to make voting Labour seem a sensible, normal part of everyday life – tell the workers you understand their resentments and hatreds, or talk to them about their working conditions?
I believe Labour’s going to have to move to the Left; anything else really is throwing the next election, if not the one after. (The fact that I have always believed that Labour should move to the Left is merely a happy coincidence.)
Next: we need to talk about Scotland.