We are living in the age of a “data” arms race. In the US “big data” was pioneered in the two Obama election campaigns. It is estimated the 2012 Obama campaign spent over $1 billion amassing and using data driven analytics. “Scraps” of data were centralised in order to understand voters. Through online transactions, supermarket purchases and social media posts a myriad of day to day online activity is logged, recorded and characteristics profiled. Via Facebook, Twitter or Reddit – if it is in the “public domain” it can be obtained. In the US political parties can now purchase data profiles of individuals in order to “target” them with campaign messages. Data can be harvested, and with the right software & resources, analysed to profile voters. This can allow political parties to create “bespoke” campaign messages, more effectively canvass and therefore GOTV (“Get out the Vote”).
Data is gathered and compiled to create complex profiles of individuals voters. Activists can be equipped with apps to feed information, concerns and contact details collected on the door-step straight back to “HQ”. The information can be combined with voter ID information from the electoral roll. This allows national campaign organisers to refine the “messages” to reflect the concerns expressed to those judged likely to be receptive to a campaign. Further, it can be seen in near real time how many doors have been knocked on, how many conversations have been had and the outcome of those conversations. For example, through data analysis a political party can more effectively understand if a certain area is “on the fence” and thus resources allocated accordingly.
At the 2015 General Election the Conservative party adopted similar polling and campaigning techniques and snapped up Jim Messina as a consultant. This was nowhere on the same scale as that in the US. “Micro-targeting” of voters, though stymied in the UK by stricter data protection laws, proved an invaluable campaigning tool. For example, the Conservatives were able to target those voters who were concerned about a Labour/SNP coalition more effectively and play upon those concerns in their election literature and leaflets. It was reported that the Conservatives spent £100,000 a month on information from Facebook. They outspent Labour on the “digital” battlefield.
By the time the next election is called there will be ever more sophisticated systems with machine learning feeding off of data collected to fuel campaigning activity. As the Conservatives did, Labour should look to the successes of Obama’s team, and more effectively make use of data engines. The Labour party needs to engage with this reality as it will become an integral part of the “electioneering” of the future. Labour needs to tools to understand how the electorate thinks, how its messages will resonate and the effect they will have. In the “marginals” the data war will be waged more fiercely than ever before. Labour needs to invest in the software for data analytics and voter profiling as part of its “arsenal” to be used alongside traditional campaigning techniques.
Even “the Donald” (President Elect Trump) engaged with data analytics. This was the centrepiece of his (shocking) electoral victory. Notably he had little “ground game” i.e the machine in place to “Get out the Vote”. To compensate for this the Trump campaign used third parties (such as Cambridge Analytics) who used software to engage data analytics in order to predict people will vote (using an estimated 5,000 pieces of information about every American of voting age) and combined this with results from personality and behavioural surveys. This essentially enabled the Trump campaign to more accurately identify voters who were most open to being persuaded and therefore focus resources. In the elections of the past the “spin doctor” was kingmaker; in the elections of the future it will be the data scientist.
Frederick Gallucci is a contributory writer for Future Labour, Vice-Chair Campaigns with Labour Retford & a masters student reading international law.