We extend our sceptical engagement with the claims made about Cambridge Analytica’s effect on recent electoral events, and share some thoughts about the liberal response to these political crises
Recently friends, colleagues, customers and even a journalist or two have asked us — did Cambridge Analytica really change the outcome an election? Was it possible for the company to use the detailed information about millions of people contained within Facebook to understand and then target Facebook users in such a way as to influence the popular vote? It is more than year since the conversations about this question began to occur and we have consistently voiced our collective opinion that the simple answer is no, Cambridge Analytica did not use the data it gleaned from its access to Facebook data to to achieve these things. Given renewed interest and a very real sense of public concern as to the nature the behaviour of both Facebook and Cambridge Analytica during the last US presidential election campaign we thought it worthwhile to set out why we reached the view we have and explore these reasons in the context of the current debate.
To put our comments in context it is useful to explain that both before and during the US elections and the UK general election of summer 2017 we have been active in developing and using techniques for finding and understanding forms of computational propaganda, we have funded our own research and cooperated with academic colleagues in a number of universities both in the UK and the US to find and understand how and why people and software tools such as Scraping, Bots and meme making are dedicated to political purposes. As part of our work we have built bots and explored and used various tools and techniques on Facebook including the purchase of advertising via the conventional Facebook approach. Similarly our consultancy practice regularly involves the application of psychology and behavioural methods in the analysis of big data in the context of commercial datasets that have nothing to do wit politics or social media. At all times we have conducted these activities within the terms of use and data access protocols set out by platforms such as Twitter, Facebook and Reddit. Following the rules has always limited what we could achieve but never so much that we thought that a project such as that conducted by Cambridge Analytica would be prevented by following the rules and this lies at the heart of our view that they didn’t do achieve what they claim to have achieved. Because we knew that even if you did have the data it still would not be sufficient to deliver what Cambridge Analytica promised.
The issue can be reduced to two separate but interconnected propositions; first that with sufficient detail it is feasible to identify who to address and how to address them in order to change opinions and actions ie voting, based upon detail information of the kind available from unqualified access to the data created by Facebook. Secondly it follows that having identified who is to be the target of an intervention contained within a Facebook post; the system would contain a sufficiently diverse portfolio of effective materials in order to place the correct post in front of the target.
The first idea which needs to be supported by evidence drawn for a source other than the claims made by Cambridge Analytica, is can people be understood in sufficient detail to give a propagandist the opportunity to change their mind on a given topic? Interestingly we have the European Union to thank for some of the most directly relevant research on this topic. The answer it seems, is that yes we can understand people in order to best select and frame arguments and evidence that might change behaviour. It is also clear from the evidence and long experience that the resources required are two-fold; you need to speak to people face to face In order to derive the information required about an the individual and to make the case for change. To put matters simply; people are different and circumstances matter. So what is required is a deep knowledge of the person and the circumstances e.g. the particular context of the issue, the moment and its place in the life of an individual. This being true it makes sense that the best results are gained when the intervention is part of a dialogue that frames the meaning of the desired action. Cambridge Analytica had access to information about people that was so impoverished as to be nugatory in this context. Following on from the question of understanding what people might need to be presented with in order to effect change is the question of what material(s) are required to act on that intelligence.
Once again the evidence from research and practice is that the form and content of material that can and typically is delivered via digital means cannot be expected to achieve the kinds of results that have been claimed for this approach. It is widely understood that not only does the material have to be carefully crafted in terms of what is know about individuals it must delivered in a variety of ways and at times and in locations that maximise its effectiveness. If it were not for the fact that the volume of research that makes this abundantly clear is manifest in a number of different disciplines from interactive marketing to town planning we would go to the trouble of listing it.
We can bring these two points together by returning to the significance of dialogue creating changes in attitude and behaviour. Obviously the training of Psychologists and Psychiatrists reflects just how difficult a proposition it is to engage successfully at a level and in a way that can lead to psychological movement. Not least because what is required is a deep understanding an empathy if you will, between two actively connected people. Technology does not yet afford a medium for replicating this experience. The dance between two people, trained or not that leads to successful dialogue uses, for example the muscles of the eyes and heart to create and extend meaning beyond written language — however it is designed and delivered. Which is not to say that in part at least this dialogue cannot be promoted between a person and a machine, because it is has been and is, on-line Cognitive Behaviour Therapy (CBT) is a good example. But this is an enormously complex and extended interaction not a glance into the news feed on Facebook
One final point is worth making, in our work we have noted as have others in different research communities, that it is possible to have an effect with the kinds of material presented by political campaigners, whether they be political parties or not. But the simple mathematics of voting systems in the UK and perhaps even more so in the USA make a material change in outcomes highly unlikely other than in the most exceptional circumstances. Most actors we are aware of know this but the marginal gains are usually deemed to be worth what are frequently seen to be marginal costs of trying.
Perhaps too this is a moment when we can address a more salient question with regard to public and private behaviour in the context of political campaigning. We have heard in every medium of public discourse as well as in private, the constant complaint, a liberal consensus if you will, that poorly educated and frequently poor white men have produced the likes of Trump and his all too pale a shadow Farage. That poor white people are the locus of blame for an outcome that could never had occurred in the face of a plurality of educated white (men). Our question is this. Why when it is rich, educated white men who are behaving with such venality, in the face of all rational evidence and common decency to trash institutions, communities and the health and well-being of so many their fellow citizens that it is the poor who are to blame? Why can it not be the men in charge, the ones actually creating the mess that makes liberals so angry should be the focus of the question why is this happening? We have seen and heard the pundits, journalists, academics and others striving to ‘understand’ why poor, ill-educated folk did this to us all. Is it because to think otherwise might raise the need to ask, to what extent is it the educated white people who are to blame for this mess? Is it because neoliberalism has served them up until now; two party systems, racist, hierarchical and divided economies were a just and sufficient way for the world to run — at least for them. We are looking forward to reading the anthropological and forensic texts that explain in painful detail why it is that rich, educated white people could do such things, over and over and over again.
We live in a world were the manners of the wealthy are the legitimate concern of the ‘public’ but it is the morals of the poor that are the meat and drink of public condemnation. Hence we would argue why it is that there is a moral panic about illegitimate action that might influence the lower ranks in society. We look forward to an enquiry as to how the graduates of our great universities, living lives of absurd material comfort come to degrade our communities in the ways they have and are doing. Perhaps then we might stop trying to lay the blame for the way the 21st century is turning out on the victims of the 20th.
