At the 2015 general election the Conservatives defied the opinion polls and unexpectedly won a 12 seat majority. In the post-election analysis, political commentators praised the party’s use of technology, particularly their highly targeted campaign to reach voters in marginal seats. But while the online world has opened up new ways for political parties to reach voters, there are also questions over how this new data is being used in local and national campaigns. New research from Dr Nick Anstead in the Department of Media and Communications finds that the use of data by political parties poses a challenge to decades-old electoral spending rules.
In British politics, two pieces of legislation are intended to limit the influence of money on the democratic process. The first, created in the 19th century, limits spending by constituency candidates in the run-up to a general election; up to £45,000 in the long campaign (six months before polling date), and up to £16,000 in the short campaign (the month before the election).
The second election spending law applies to political parties at the national level and sets much higher spending limits for their election campaign. The limit is calculated by multiplying the number of constituencies being contested by a party by £30,000; if for example, a party contested all 650 seats in the UK, they could spend a maximum of £19.5M in the year before an election.
The existence of these two different laws setting out separate spending limits – one for local campaigns and one for national campaigns – is a source of potential confusion, or even abuse, if a party’s central resources are used at the local level. For Dr Anstead, one of the stories of the 2015 general election was how data organised and held at the national level was fed into direct personalised campaigning techniques at the local constituency level. He conducted a series of interviews with representatives from the major parties to understand how this increasing use of data and technology is influencing the political process.
“If you have a Facebook advert that addresses issues that resonate locally, where should the spending be attributed? Similarly, what happens when door knocking is organised by a national dataset, but used to target individuals who might be willing to switch their votes in swing seats? These technologies and communications tools blur the lines between national and local”, Dr Anstead says.
Arguably, the two tiers of electoral laws have been outpaced by technology in the modern political era, with regulators unable to act fast enough to catch and punish transgressors. In March 2017, the Conservative party was fined £70,000 following an investigation into their election campaign expenses by the Electoral Commission, while criminal charges are being pursued against a Conservative candidate and members of his team over his 2015 election expenses. But the election had already been won two years earlier.
A further issue data-driven campaigning techniques raise is whether technology reduces democratic plurality. Smaller parties do not have the mighty resources of Labour and the Conservatives, and therefore may find it even more difficult to compete. The term for this development is cartelisation— when the obstacles are so great that they consolidate the power of the already powerful.
“We used to have an idea about the internet that there are very low-barriers to entry. While there are examples where politics is practiced like this, such as single-issue groups or student politics, national politics is different,” Dr Anstead says.
“If you want to pursue swing voters in in a complex political environment, it requires sophisticated methods, and it is expensive. This means it is out of reach of most parties,” he adds.
Following a bruising experience for the Labour party in 2015, Dr Anstead observes that the party fought a different campaign in the 2017 snap general election, winning back a number of seats from the Conservatives and forcing a hung parliament. “Labour are now being credited for running an organic campaign, with a big role for activists, in 2017. This is the exact opposite of why the Conservatives were thought to have won in 2015, where they fought a disciplined and organised operation,” he says.
For Dr Anstead, the truth is probably somewhere in between. “Data-driven targeting can be very effective, but having an effective organic campaign is also useful. But we are learning that Labour also spent significant sums of money on targeting through Facebook adverts in 2017, something that the Conservatives did successfully in 2015,” he says.
While the mechanics of technology-driven campaigning are an important component of campaigning, political success still hinges on an effective central strategy. “In 2015 the Conservatives had a very powerful grand narrative, talking up the threat of an unstable coalition between Labour and the Scottish National Party, and focusing on their claims of responsible management of the economy. But in 2017, Labour’s anti-austerity message seemed to connect with some voters,” Dr Anstead says.
These ‘grand narratives’ intersected with each party’s data-driven, localised campaigns, with targeted policies from Labour, such as the abolition of tuition fees, apparently mobilising the party’s youth vote and increasing the overall turnout.
This unexpected development affected the accuracy of the election data held by both parties. Dr Anstead says: “With any data-based strategy, it’s only as good as the ability to predict the turnout. The turnout patterns changed dramatically in 2017, and this meant that young voters were not factored in.
“If you don’t reach voters or ignore them, then the data models become unreliable. When we are analysing elections, it is important to acknowledge the importance of every part of the campaign and how they relate to each other.”