How Women in Political Data plans to close the polling gender gap

There are plenty of low-paid women, but economic divisions are still based on male-dominated jobs. 

NS

Sign Up

Get the New Statesman's Morning Call email.

This month, Women in Political Data launched at YouGov. Open to all women interested in political research and analysis, WPD has been established to address a disparity that’s often overlooked despite considerable implications for our politics: the polling gender gap.

Polling has the ability to shape narratives and opinions. It is clear that our polling industry is itself a key player in our politics, but its composition is given significantly less attention than that of politicians.

In very different ways, the 2015 and 2017 elections were shaped by pollsters. Each of these elections were centred on debates and beliefs dictated by what the polling appeared to be telling us. The debate in 2015 was shaped hugely by the Scottish National Party propping up Ed Miliband, because the polling appeared to say no-one could form a majority government. In 2017, pollsters underestimated Jeremy Corbyn, and the debate was centred on how a supersized Tory majority would change the country. The EU referendum, on the night, was also a fall-off-your-chair event again – because of a false narrative.

Not only did voting intention polling directly affect the elections and referendum, the context and aftermath, too, were shaped by pollsters. The context of the 2015 election and the EU referendum were very much defined by narratives on what the data told us about “left behind” Britain. These were the pinnacle of a growing working-class anger, one that swept an elite off their feet. Yet the way we came to define class was very much on the preposition of the Angry White Man.

Men often dominate discussion in focus groups, competing with each other to be the loudest, and tend to make for eye-catching stories, particularly based on the Angry White Man fighting the system. They then make for the headline act of reports and articles, of sociological long-read tributes to an era defined by the anger of the BNP, UKIP and Trumpism.

Ian Warren, who has helped develop Women in Political Data, told me the challenges of focus groups that led him to split his by gender: “Some men want to impress each other in focus groups whilst women don’t pretend to try. It means it’s easier to draw insights from women, almost all of which cut through complex subject matter with devastating honesty. I often come away from focus groups thinking that women should just run everything. I’ve sometimes felt the same about the internal decision-making structures of our political parties too. “

The implications of class and gender particularly is a historical issue that also extends beyond the anger of our own era. Social class has dominated the debate in Britain, but schema for class defaults to a male bias. Class has generally been measured via heads of households, which presents a historical bias. Also, the old divide of manual versus non-manual – one that continues to dominate debate including the way we discuss the metropolitan versus rural and industrial – is presupposed on male jobs. Women in the public sector can be low-paid, but narratives have historically barely registered their existence. In the present day, discussions about what technology and artificial intelligence mean for the world of work are more likely to be hinged on the manufacture of a car than the livelihood of a cashier, for instance, with predictable implications for gender.

Gender as a prerequisite for voting behaviour is also overlooked itself. Women may express politics differently – they may express anger differently, and consequently express this anger often by voting differently. Survation recently found that today’s Labour lead is driven by female voters. Labour enjoys a 16 per cent lead among women, while the Conservatives lead among men by a whisker. The left-wing commentator Owen Jones has pointed out that the narrative within the Labour party wilfully ignored the drive of female voters. The story of the female voter is, presumably, less dramatic than that of the loudest man in the focus group.

Women will, like men, harbour prejudices and biases that determine political opinion, but it is clear that gender is a prerequisite to how we express these prejudices at the ballot box. Ignoring how gender drives voting could have similar ramifications for polling as underestimating young people did in 2017.

Polling and research shapes how we view the electorate and a culture, but it is all too often conducted with masculinity in mind, with the inevitable outcome being masculine narratives.

A savvy researcher will try to compensate for the biases that lead to the dominance of these narratives, but this isn’t a given. It is certainly not a given if there are few female perspectives in the room or industry as a whole.

Polling and research companies are headed overwhelmingly by men. The CEOs of almost all the major UK polling companies are men. Research teams are more diversified, but the public face of the industry defaults to male. This leads also to the danger of public voices in debate being overwhelmingly male too. All-male panels still haunt conferences, events and TV. The composition of the polling industry shapes our political commentary.

The conduct and outcomes of political research being so male is in spite of social sciences enjoying a gender balance at university. Yet statistical analysis is a Stem subject within the social sciences, and shares the same characteristic of gender disparity as elsewhere with at least equal implications – but with less attention.

For coding, there has been a surge in organisations encouraging women into tech, with fantastic results. There, there was also an acknowledgement that gender disparity affected the way technology shaped our lives. Women in Political Data acknowledges the way in which gender disparity affects our political lives, and seeks to confront it by building the political data analysis and research skills of a new generation of women.

Women in Political Data has begun this effort with an event in YouGov, with a panel of female leaders in the industry. Participants can join the online community where data will be freely shared and they will be guided on how to use it for research, as well as get their research published in the press.

The organisation will hopefully not only inspire more women into the political research industry, but also kick-start a long overdue conversation about the implications of the polling gender gap to our politics.