Who Will Win in 2015? Peter Kellner Talks to Birkbeck

By Dr Benjamin Worthy

Peter Kellner, expert pollster and President of YouGov, spoke to the Birkbeck Centre for British Politics and Public Life on Wednesday 5 November. A podcast of the talk is also available.

Peter spoke of how influential polls could be. He gave the example of the YouGov poll run by the Sun in August 2013 before proposed military intervention in Syria in 2013. This polling had a real impact on the subsequent debate and may have contributed to the narrow defeat of the vote on military action (or to put it more precisely, on the government motion).

Public opinion can also be fickle – see the changes in public opinion over the War in Iraq and the fluctuation in the ‘support’ and ‘oppose’ column between 2002 and 2007. The public can also get it wrong (see how mistaken we are about everything here). Peter spoke about the need for leadership and the fact that a leader’s job is to sometimes to tell people they are wrong. Immigration is good example – see this gap between perceptions and reality.

So how about the big question – who will win in 2015? In brief, it isn’t clear. Most elections are decided not by switches to Labour-Conservative but by undecided and Liberal-Democrat voters. However, for 2015 there is not one but three wildcards.

Wildcard 1: How will the Liberal Democrats do? We do not know whether or to what extent Liberal Democrats will suffer (or not) for being in government. Previous election results were based on Liberal Democrats as a ’third party’ and a ‘protest vote’. How many seats will they lose from their 57? Will they be down to 30? 20? Or will their famously efficient ground organisation machine save them? This analysisconcludes ‘there are so many possibilities, you can make up your own mind what it all means’.

Wildcard 2: How will UKIP do? This is less about which seats they may capture – possibly 10 but more likely four to six. More importantly, how may Labour versus Conservative seats will they throw in a particular direction? Here the number may be many more (see this blog by our own Eric Kaufmann and this analysis of UKIP support).

Wildcard 3: How will the Scottish National Party do? A recent YouGov poll gave the SNP an astonishing 19 point lead in Scotland, enough to capture 31 seats from Labour. Even if this does not happen, the SNP could capture enough of them to deprive Ed Miliband of victory. This is indeed Labour’s Scottish nightmare.

So these three wildcards may well shape who wins or loses, without mentioning even more complications such as the Greens, now polling higher than the Liberal-Democrats. The most likely result is some sort of ‘messy coalition’ made up of various parties of one combination or another. One thing is sure, as Peter puts it here, ‘Those days of decisive, first-past-the-post election outcomes might be over, at least for the time being’.

Rich man, poor man, politician man

The question of voters’ reactions to the personal wealth of politicians is increasingly topical. Even in the United States, where money is generally equated with success, Mitt Romney’s vast fortune (and low tax rate) was used against him in the Republican primaries. In Britain there has been a public debate as to whether politicians should have to publish their tax returns – with candidates for the London Mayoralty (mostly) doing so.  The issue is certainly live, but at present academic research can contribute little to the debate.

To investigate the public’s reactions to wealthy (or not so wealthy) candidates we ran an internet survey experiment where respondents were asked to rate and choose between two hypothetical candidates based on short biographies. We applied two experimental manipulations, one to wealth and the other to occupation.  Splitting the samples, we varied occupation between a self-made businessman and employee of an international finance company, and we also varied the amount that they earned.

The effects – with the headlines reported here – were powerful.  We found that voters preferred the self-made businessman to the financier, but that regardless of occupation they reacted negatively to financial success.  As we increased the amount earned by our hypothetical candidate, so their popularity declined.  Voters did not reward candidates for their financial success, rather they turned against them.

Yet what was also revealing was that voters did not respond homogenously to our experiments.  We found notable differences by both social class and party support.

The experiment used a split sample technique, showing almost identical candidate profiles to respondents but making slight alterations in wording.  It compared two hypothetical candidates – John and George – and measured the differences caused by making slight changes in John’s profile.  We measured three perceived candidate characteristics – approachability, experience, and effectiveness – as well as asking which candidate respondents preferred overall, but for reasons of space, we focus here just on the overall preference figures.

The top half of first table (below) shows the difference in the way respondents chose the candidates when we presented them with John as a businessman who earned some £28,000 (the average male salary).  As is clear, John was preferred over George (who was a solicitor earning £45,000 per year) amongst all groups but especially strongly amongst those in the C2 and DE social groups.

The bottom half of the same table shows the results when we increased John’s income to a cool one million (but kept every other aspect of the profiles the same).  Now George was preferred to John amongst every social class.  One way to look at this is to examine the ‘cost’ of the increase in income – by looking at the change in John’s lead from when he earned £28k to when it was a million.  Amongst ABs, for example, the lead goes from +27 to -19, a change of 46 percentage points.  Amongst C1s, the difference is 54 points. But amongst C2s and DEs, it is 65 points. So everyone is put off by wealthy candidates, but the working class are put off more.  We found exactly the same when we repeated the experiment but with John as a financier instead of a businessman: the total cost of the increased income was 49 points amongst ABs, but 67 amongst DEs.

John’s occupation/income

Social class

John

George

Difference

Businessman/28k

AB

51

24

+27

C1

44

22

+22

C2

57

13

+44

DE

58

14

+44

Businessman/1 million

AB

25

44

-19

C1

18

50

-32

C2

28

49

-21

DE

26

47

-21

The differences by party support were even more striking, as shown in the table below.  Again, as levels of income rose, so John became less attractive to supporters of all three main political parties – but at a very different rate.  Labour voters were especially turned off by the rising income: from favouring John by some 41 points when he earned £28k, if John earned a million they favoured George by 52 points – a massive 93 point transformation.  Lib Dem voters had a similar experience, favouring John by 41 points when he earned £28k but with that transforming into a 26 point deficit once his income increased to a million.

John’s occupation/income

Party

John

George

Difference

Businessman/28k

L

59

18

+41

C

43

25

+18

LD

58

14

+44

Businessman/1 million

L

12

64

-52

C

38

36

+2

LD

23

49

-26

Conservative voters, however, reacted differently.  The total cost of John’s extra income amongst Conservative voters was a still significant 16 points, but even after John was earning a million per year they still preferred (albeit just) John to George.  (We also showed some respondents a version in which John earned £100,000 – and Conservative voters also preferred that John to George, this time by three points).  There clearly was, therefore, some Conservative respondents who saw John’s financial success as positive and a reason to want him as their representative.

However, when we changed John’s occupation to that of a financier rather than a businessman, we did not find the same effect.  Conservative respondents were still relatively more positive towards John than George, but the cost of John’s extra income increased to some 37 points even amongst Conservatives, and they no longer preferred John to George.  In other words, Conservative-inclined respondents were less put off by candidates with higher levels of income and they made a sharper distinction between money earned by someone who had set up their own business and that earned from a multi-national company than did other respondents.

So wealth could be a negative factor for all candidates, but it is likely to be an especial negative for Labour and Lib Dem candidates.

Rosie Campbell is Senior Lecturer in Politics at Birkbeck, University of London; Philip Cowley is Professor of Parliamentary Government at the University of Nottingham.

This post originally appeared on Ballots & Bullets, the blog of School of Politics and International Relations, University of Nottingham.

NOTE: Fieldwork using the YouGov Plc UK panel (350,000+ adults who have agreed to take part in such surveys) was undertaken on 10-11 April 2012 (sample of 1,727 people) and 11-12 April 2012 (sample of 1,686 people). Figures have been weighted to be representative of all UK adults (aged 18+). All text is solely the opinion of the authors.

Men’s voting behaviour: it’s a hunter-gatherer thing apparently!

By Dr Rosie Campbell & Dr Jason Edwards

BBC Radio Four’s Beyond Westminster recently ran a programme on ‘The Men’s Vote’. The programme asked why politicians and the media seem more interested in targeting women’s votes than men’s. Rosie Campbell, Peter Kellner and Hopi Sen debated the issue.

Peter Kellner claimed that we can look back to early humans to help explain differences in men and women’s political attitudes and behaviour. He said “I think there is, probably, still in the different gender DNAs something that goes back to the early human, men being hunter/gatherers and the women being the home-makers” and that “hunter/gatherers are risk takers – have to be – home-makers are risk avoiders – have to be”. Peter viewed these pre-historic differences in gender roles as pivotal to understanding how men and women relate to politics today. Hopi Sen and Rosie Campbell were quick to disagree.

So what’s wrong with using hunter-gatherers to explain gender differences in voting behaviour in modern Britain?

The idea that there are ‘gender DNAs’ is, to say the least, strange. There is no such thing as male or female DNA. The DNA of all human beings is virtually identical. The only difference between men and women in the organisation of DNA within the cell is that women possess a pair of X chromosomes whereas men possess only one X chromosome in addition to a Y chromosome. That chromosomal variation is what explains sexual dimorphism in the human species – the different sexual organs of men and women and a different balance of hormones associated with their roles in sexual reproduction. So presumably, what Kellner means is that there is a biological basis (genetic, chromosomal, hormonal) to the supposedly contrasting social behaviour of men and women and that this ‘natural’ behaviour is expressed in their ‘natural’ condition, i.e. the environment in which early humans (‘hunter-gatherers’) found themselves in the Pleistocene epoch.

Oddly, Kellner claims that men in this environment are the hunter-gatherers wheareas women are home-makers. But most paleo-anthropologists agree that women did (and do) the majority of gathering in hunter-gatherer societies, and they also argue that there was (and remains) no absolute sexual division of labour: in most bands, men also gather and women also hunt. Of course, hunter-gatherer bands are also, for the most part, nomadic, so it’s a bit weird to say that women were ‘homemakers’ in societies that don’t have ‘homes’, at least as fixed abodes in which children are nurtured.

Another faulty assumption that Kellner makes is that hunting is a risky, and economically key, activity only suited to men who are (hormonally) equipped to be risk-takers. Depending on the environment, hunting can be very low-risk and economically marginal. In times of abundance, the Indians of the Pacific North-West could stand at the side of the river and catch leaping salmon in their arms. At the historic peak of the buffalo hunt in the 19th century, the Plains Indians, on horseback and armed with rifles, would kill so many bison that they took the choice parts – the pelt for clothing, the eyes and liver as delicacies – and left the carcasses to rot in the sun.

The point is that Kellner repeats an old chestnut: in their ‘natural’ condition, the social roles of men and women are distinct and determined by biology. This is to vastly underestimate the complexity and diversity of ‘primitive’ societies. In fact, the social roles of men and women in traditional ‘hunter-gatherer’ societies are as much determined by the material environment, by their religious and magical practices, by the specific forms of social and economic relations they possess, as they are by biology. That, of course, does not amount to saying that biology is unimportant, but it has always been mediated by the social, economic and cultural structures of human societies which are, to an important extent, autonomous of biological determination.

The mistake Kellner makes with respect to biology and gender is compounded by the evidence from contemporary politics. First there’s more similarity than difference in men and women’s political preferences in contemporary Britain. In the post WW2 period women were more likely to support the Conservative party than men, but this tendency has gradually declined, so that at most elections women are only significantly more Conservative in the very oldest generations, and even then the percentage point differences are usually only in single figures.

The British gender gap 1964- 2010 (Adapted from Norris 1999)

(The gender gap is calculated as the difference between the Con-Lab lead for women and men. Gallup polls 1945-59; BES 1964-2010)

Second, where there are differences in attitudes they seem to shift across the life cycle. Women tend to be more concerned about education and healthcare than men, but this isn’t stable across age groups; younger women are usually more concerned about education and older women more concerned about healthcare.

Third, where there are gender gaps in political attitudes the percentage point differences are not startling. According to the 2010 British Election Study post-election face-to-face survey twice as many women as men said that the NHS was the most important issue facing Britain, but only 2.6% of women and 1.2% of men selected the NHS as the most important issue. The issue that most voters selected as most important was, not surprisingly, the economy (42% of men and 34.1% of women selected the state of the economy); an eight percentage point difference is worthy of note but does not seem to indicate a seismic difference in the way men and women prioritise political issues. Small gender variations in political attitudes, where women are slightly more concerned about social services and men worry more about the economy and taxation, are regularly evident in Western democracies, but they rarely reach double digit percentage points. It seems most likely that these small distinctions reflect the vestiges of traditional gender roles, where women have been more associated with the family and caring responsibilities and men seen as providers. But crucially, as women have moved into employment and higher education differences in political attitudes and behaviour between men and women have declined.

These trends undermine Kellner’s hunter-gatherer thesis. If gender differences in political attitudes are hard-wired into our brains as part of an evolutionary process that started with the earliest humans, surely we would expect to see large and predictable differences between men and women that are relatively stable over time and place?

This debate is illustrative of the ongoing argument in the social sciences, and beyond, about gender difference. It’s the age old nature/nurture debate that tore second wave feminism apart- when some radical feminists claimed that men were essentially aggressive and couldn’t be reformed and broke off from liberal feminists- many of whom shared their beds and their lives with men who they believed to be pretty ok.

So why are so many political commentators tempted to take us back to prehistory? Perhaps it’s all in the DNA . . .