Tuesday, 6 October 2015

Can Corbyn win? - A Look at the Stats

Oisin Gilmore: For over the last 100 years or so progressive politics in Western Europe has been dominated by a small number of major parties. Some of them such as the Sozialdemokratische Partei in Germany, the Labour Party in the UK and the Parti Socialiste in France continue to play this leading role, while others are now just history such as the Parti Communiste Francais, which has shrunk to insignificance, and Partito Comunista Italiano, which has transformed into the Democratic Party. Therefore, the development of these parties has not been merely of domestic interest, but rather it has been of interest for the left across Europe.

It mattered for everyone when the Mitterand government failed in 1983, it mattered for everyone when Lafontaine resigned from the SPD and it mattered for everyone when the Blairite’s launched the ‘Blair Revolution’.

After 30 or so years of shifts to the right in each of these parties, this summer saw a distinctive leftward shift in the UK Labour Party.  Against all odds, Jeremy Corbyn, a rank outsider, won the leadership election. How Corbyn came from the miniscule Socialist Campaign Group (which prior to his election contained 9 of the 252 Labour MPs) to win the leadership election is a story for elsewhere. The fact is that he won. What he can do with that victory is today a more pressing question. And what he can do with his victory depends ultimately on two issues: the policies he proposes and his ability to get those policies implemented.

Here I want to look at this latter question, and I want to focus on it in a very restricted way by looking solely at his ability to get elected.

During the election campaign for leader it was often said that Corbyn’s unelectable and some have joked and others worried that, by electing Corbyn, Labour has lost the 2020 election in 2015. Given the apparent negative impact the perception of Ed Miliband as an uncharismatic, disconnected Westminster insider had on Labour’s fortunes, this argument curiously skates over the question of whether the three more mainstream candidates were capable of leading Labour to electoral success.

But really, the argument isn’t about Labour’s ability to win an election. The argument is much simpler. For many, Corbyn is simply too left wing.

In this blog post I want to ignore the question of whether or not Corbyn is too left-wing, in fact I will ignore completely his policies and consider in isolation the quantitative question of: Can he get enough votes to win in 2020.

2015 was a major failure
The first thing to understand about where the Labour Party stands after 2015 is to how large the electoral failures were in both 2015 and 2010. Before the 2015 election many Labour supporters expected a Labour victory, at worst they expected a victory for a Labour led coalition. They thought it was a small job. But building on the disaster of 2010 was no easy task. And it was a task Labour failed, winning only 232 seats, 94 seats short of a majority.

In 2010 less than 19% of the electorate [I am considering the electorate here, not voters i.e. I am counting non-voters.] voted for Labour, less than in the famous electoral disaster of 1983. In 2015 this scarcely increased and Labour won the votes of 20.1% of the electorate, the exact same as in 1983. (See below.) [Data compiled from wikipedia UK election articles.]

The reason why Labour performed electorally better in both 2010 and 2015, when it won 258 and 232 seats respectively, than in 1983, when it won 209 seats, was due primarily due to two factors: firstly, the Tories are also performing much worse than they did in the 80s and, secondly, the level of voter abstention is much higher. Fewer people have voted in each parliamentary election since 2001 than in any election between 1974 and 2001.

 In both 2010 and in 2015, Labour electoral support was reduced to its traditional heartland. Indeed, it has actually been reduced electoral dependence on the socio-economic impact of what is in the land. As the BBC Wales journalist Vaughan Roderick pointed out, Labour seats are concentrated where there used to be coalfields.

This may appear to be merely a curious statistical artefact. Leo Hickman pointed out that the results also maps onto life expectancy, whereby Labour seats map pretty neatly onto the areas in the UK with the lowest life expectancy.

However Dr Alan Fernihough of Queens University, Belfast has performed a simple statistical test of this and even after controlling for longitude and latitude, he finds “Being located on or in close proximity to a coal field reduces the Tory vote share by about 20%.”

All of this should be quite worrying for Labour. An extreme but not unfounded way of stating this problem is that it appears Labour today are only able to get people to vote for them because their grandfathers were coal-miners.

While the Labour HQ worries about the impact of Corbyn’s victory on Labour electability, it would appear that Labour has much more to worry about electorally than just Jeremy Corbyn.

Corbyn’s Strategy
In response to this Corbyn has been explicit in what his electoral strategy is. He has said he wants to focus on “Young people who didn’t register, who didn’t vote.” He argues that “those that did vote were overwhelmingly Labour, so I think there’s a whole area there, and this campaign is demonstrating that. Young people, who are very political, but are not interested in party politics, coming forward. 

Secondly, the numbers of reliable Labour voters who disappeared into the arms of UKIP or non-voting because they didn’t feel the Labour Party represented anything they wanted to hear, do or say. I think we can grow our support that way. Do we have to win back people who voted for other parties? Yeah, but we have to say to people, in a very clear way, what we’re offering.”

His strategy is to focus on young voters, non-voters and voters for people Labour has lost to other parties, in particular UKIP.

This may strategy to focus on non-voters may have some sense.

Between 1955 and 1997 between roughly 22% and 27% of the electorate didn’t vote. But since 2001 roughly between 33% and 40% of voters have stayed away from the polls. Exactly what drove this drop in turnout isn’t clear but a study by Ian McAllister of the Australian National University found that much of the non-voting in 2001 was associated with low income, welfare dependency, rented accommodation, being a member of an ethnic minority or resident  where the Labour majority was very high in 1997.

Further, Labour performed best among younger voters who are also the least likely to vote. (See below.) [Data taken from Ipsos-MORI]

It is perhaps worth mentioning here that some in Labour have expressed strong doubts about this strategy by pointing to polling data collected by the Trade Union Congress that appears to show non-voters not expressing traditionally right-wing concerns about Labour. (See below.)

However, people should be very cautious in interpreting this data because, as can be seen, no traditionally left-wing concerns were given as possible answers by the pollsters. Its unwise to conclude that people didn’t vote for Labour because Labour were too leftwing when the only options given to them in a survey is that either Labour were too leftwing or “Other”!

Of more serious concern for Corbyn’s strategy is the “The mountain to climb: Labour’s 2020 challenge” study produced by the Fabian Society which states that after taking into account boundary changes Labour will need to win not 94 seats, but 106 seats in 2020. The study then considers the swings needed in each constituency if this is to be achieved. They find that for an outright UK wide majority, a swing of 9.5% would be required.

In other words Labour would need to win 40% of votes, up from 30.4% in 2015. This is no small task. They argue that due to the decline of the Lib Dems, 4 out of 5 of these votes will need to come from Conservative voters. But, in this study they appear to ignore the issue of non-voters.

More recently Olivia Bailey, Research Director at the Fabian Society, explicitly warned that “Under Corbyn’s electoral plan, prospects for victory look bleak”. Pointing to the boost in turnout in Scotland in 2015, which benefitted the SNP, she warned that “even if Labour secured a Scotland-style boost in turnout of 7.3 per cent across the English and Welsh marginals, a maximum of 52 seats could be won, if each new voter backed Labour – still dozens of seats away from a majority.”

Overall the evidence here seems ambiguous. It would appear that without winning back non-voters Labour would need to win over an inconceivable number of conservative votes when the number of Tory voters is already at a historically low point. And the evidence would appear to support Corbyn’s belief that getting non-voters and young non-voters in particular to go to the polls will lead to Labour votes.

However, although it may well be the case that Corbyn was, of the four leadership candidates, the most likely to be able to inspire non-voters to go to the polls, the increase in turnout required for a Labour victory appears to be very high. And regarding his other strategy of winning back UKIP voters, things look even less hopeful.

The problem with Corbyn’s  concern for the “numbers of reliable Labour voters who disappeared into the arms of UKIP” is that this number is simply much lower than the media would lead you to believe. Whilst pundits often claim that much of UKIPs support comes from working class former Labour voters, the evidence strongly suggests otherwise. While it is true that much of UKIPs support comes from voters with lower incomes and lower education levels, very little of UKIPs support comes from former Labour voters.

Regarding the average UKIP voter some data comes from the not particularly reliable but frequently reported YouGov data. (See below.)

More reliable data on where UKIP’s votes are coming from comes from the British Election Study.

The data on voter flows from one party to another show clearly that only a small proportion of the increase in support for UKIP in 2015 came from Labour voters. (See below.)

A much larger proportion of UKIP support in 2015 came from former Tories and from former LibDem voters. This is not a 2015 specific finding. Prior to the election Geoff Evans and Jon Mellon of the University of Oxford reported similar finding for the period leading up to 2014. The below diagram shows voter flows between parties from 2015-2010 and voter flows to UKIP for the 2014 local and European elections.

As can be seen, very few voters went from Labour to UKIP in either 2005-2010, 2010-2014 or 2010-2015. Quite simply, even if Corbyn wins back the “reliable Labour voters who disappeared into the arms of UKIP”, he’ll win back very few voters, certainly not enough for it to have a major electoral impact.

One note of caution here about my scepticism regarding Corbyn’s UKIP strategy is that the above data discusses voting patterns at an aggregate level and of course in the UK national vote shares don’t matter much. What matters are votes at a constituency level. And if we look at where UKIP gets its largest votes, these often do map onto where Labour won in 1997 but lost in 2015.

 For example, pretty much the entire Thames estuary went Labour in 1997, but went almost entirely Tory in 2015. As can be seen in the above diagram, this is one of the places where the UKIP vote is strongest. [Taken from Benjamin Hennig of the University of Oxford’s blog ]

So while it may not be the case that a lot of Labour votes went to UKIP, it may be the case that important votes went to UKIP and/or that those votes were lost for the same reasons that Labour lost votes to the Tories or to abstention. But this is very speculative and intended only to point to how my scepticism may possibly be unjustified.

As this article is for an economics blog I wanted it to be primarily quantitative and avoid discussing the political aspects of Corbyn’s electoral prospects too much. However, to conclude I want to briefly discuss some of the implications of all of the above for Labour’s electoral strategy.

During the election for Corbyn, the Fabians claimed that in order to win the next election 4 out of 5 votes would need to come from former Tory voters. This claim appears to have been based on the assumption that turnout would not increase. As a strategy it is hard to believe that this would have worked regardless of who led Labour. In both the 2010 and 2015 elections the Tories won roughly 24% of the electorate.

In every other election they won since 1923 , they won with over 30% of the electorate [In 1923 the vote was quite evenly divided three ways between the Tories, the ascendant Labour Party and the newly reunited Liberals]. Or in terms of voters not the electorate, in 2010 and 2015, they won roughly 36% of the vote, every previous electoral victory required more than 42%.

If the Fabians are correct and a swing of 9.5% is required for Labour to win and if aimed to get 4 out of 5 of these from Tory votes, they would need to reduce the Tory vote to 28.4% of voters. Quite simply, regardless of who won the leadership, this was almost certainly never going to happen. For Labour to win in 2020 it needs to focus strongly on getting non-voters to vote. This is Corbyn’s stated strategy and given his ability to inspire people to join the Labour Party in their tens of thousands it is likely that of the four candidates Labour chose the candidate most capable of inspiring non-voters to vote.

However, it is hard to believe that this electoral strategy will work by itself either. A 9.5% increase in turnout is unlikely to happen. Even if he wins some of the Green’s 3.8%, or some of the SNP’s 4.7%, it will be hard for him to make up the numbers without winning over votes from the Tories or UKIP. And to reemphasise the point, UKIP voters are by enlarge right-wing former Tory voters, not former Labour voters.

Corbyn therefore appears to be stuck between a rock and a hard place. In order to inspire turnout and win back votes from the Greens and SNP he will need to continue to appeal to the left, but in order to get the swing he needs he will need to also appeal to the right. While it is almost certain that appealing primarily to the right as Liz Kendall advocated would have been a complete electoral disaster, and while it is quite possible that Corbyn will be able to substantially increase Labour’s vote in 2020, it is hard to believe that there is a strategy by which he or any other Labour leader could win.

The implications of Corbyn’s leadership of Labour will be of major significance for the left across Europe. In order to win in 2020, he will need to dramatically shift political discourse in Britain over the next five years. Regardless of the leader, Labour can’t depend on a few swing voters.

If against these difficult odds he actually manages to win, it will be of major significance for European social democracy. It is not hard to imagine such a victory inspiring similar leftward shifts in the Labour Party’s sister parties across Europe. What happens in the UK is therefore worth watching for all of us.

Oisín Gilmore is a PhD candidate in economics at the University of Groningen


Paul Hunt said...

When the British Labour party changed the system to elect its leader to give its MPs the same weight as members, TU affiliates and £3 supporters and when these members, affilliates and supporters elected Jeremy Corbyn they effectively, as a self-selecting selectorate, over-rode the democratic legitimacy of the elected MPs who did not support Jeremy Corbyn. However, despite its flaws, the UK remains a representaive parliamentary democracy. Unfortunately, while Jeremy Corbyn relies on this selectorate (and hopes to increase it) to enforce his leadership mandate and his policy proposals he will take the Labour party outside the parliamentary process and it will become a protest movement.

Paul Hunt said...

This is an excellent post. However, the members, TU affiliates and £3 supporters who gave Jeremy Corbyn his stonking majority are rejecting sensible elements of the programme on which the addmittedly reduced number of Labour MPs were elected. The Tories are cheerfully stealing a number of these sensible proposals - a living wage, some fiscal devolution to local authorities and an infrastructue commission - to capture more of what they describe as the "common ground". In response Team Corbyn is seeking to shift the terms of the economic policy debate far beyond what a majority of British citizens will consent to - even if elements of what are being proposed resonate with them.