by Anthony Painter
Sometimes you find yourself picking up a book with a degree of scepticism and ambivalence. This was one such occasion. Issenberg’s pacy Moneyball-style look at the evolving statistical and psychological science of winning campaigns had potential for ‘overstatement of case’ written all over it. It didn’t help matters when the final sentence of the introduction argued that this new application of science had enabled campaigns to ‘start treating voters like people again.’
And yet, I ended up convinced by Issenberg’s argument against my own expectations. The techniques and approaches he outlines within the Victory Lab, if applied with imagination, have the potential to re-engage millions in democracy – up to a point. His core contention is that voting is a behavioural act. As he eloquently puts it:
“What if voting wasn’t only a political act, but a social one that took place in a liminal space between the public and private that had never been well-defined to citizens? What if toying with those expectations was key to turning a person into a voter? What if elections were less about shaping people’s opinions than changing their behaviours.”
There is an old rational choice puzzle that all undergraduate students of political science are presented with. Why do people vote when, from an instrumental point of view with a low chance of influencing the outcome, it’s an irrational act? Rationalistic models of human behaviour have shown themselves in economics, politics and in psychology to be completely inadequate. Issenberg reaches instead for the behavioural science of the likes of Daniel Kahneman, Amos Tversky, Robert Cialdini, and Richard Thaler – all names that will be familiar to those who have been following these debates.
The answer to the rational choice puzzle is, of course, because human beings aren’t just instrumental calculating machines. Any (political) science that can’t cope with that insight is for the birds.
Essentially, the Victory lab is the story of how a group of US political operatives – in both parties and beyond – have applied these insights to political campaigns. How do you persuade people to vote for a certain candidate and how do you persuade them to go out to vote? Beginning with Harold Foote Goswell in the 1920s, political scientists have been studying these questions in an experimental sense. In 1927, Goswell published a book called Getting out the vote. The language of campaigns has a longer lineage than may be assumed.
The innovation of randomised control trials where different techniques of persuasion are sampled with different groups and the application of algorithmic models to voting probabilities were later innovations. These techniques have been increasingly applied within campaigns over the last decade or so. In the battle between the ‘geeks’ and the ‘gurus’, it is the former who have increasingly won. They can say what works better – helpful when it comes to spending scarce campaign dollars.
Much was made of the application of ‘big data’ to Obama’s victory last year. Perhaps it was this over-emphasis that drove my mild scepticism. The sabermetrics that lay behind the Moneyball success of Billy Beane’s Oakland As seem to have diminished utility over time: others have adopted the technique. So it is or will be with campaigns. The statistical technology will spread. Everyone will do it. I very much doubt that the Romney campaign missed off this technique- its problems were far more profound than that. It has or will become a necessary but insufficient campaigning tool in a short space of time.
In the UK, the application of ‘big data’ and micro-targeting has barely begun. Those Lord Ashcroft polls that he very kindly publishes show that the Tories are doing it. Underneath the top-level, voting models he publishes is a wealth of geographic, attitudinal, social and demographic data that is being scoured over as we speak. It would not be surprising at all if the Conservatives are already running randomised control trials on various persuasion media and messages. This will be through direct mail so below the radar.
One lesson from the US experience is that this innovation occurs outside the party machine. In unions, campaign groups, new institutions such as the Catalist data company, America Coming Together, Yale university and the Analyst Institute, “Bell Labs” style innovation can occur without cultural resistance. These innovations eventually found their way into the Obama campaign and a series of Republican campaigns – not least the Bush re-election campaign in 2004. For Labour to adopt these techniques effectively, it may have to create an independent Lockheed Martin style “skunk works2 – immune from day-to-day interference for a while.
In the context of a campaign, focusing on people’s needs and motivations is a positive thing if the objective is to engage them in the democratic process. The major experience of micro-targeting in the UK was Labour’s 2005 campaign which was a mess. Mark Penn, who also advised Hillary Clinton in her unsuccessful bid for the Democratic presidential nomination, advised the application of micro-targeting in an above-the-line fashion. Jamie Oliver’s school dinners initiative was appealing to a small segment of the population – ‘Soccer mums’ so-called in the US. For everyone else, Labour’s campaign was just a confusing cacophony. These techniques should be below-the-line rather than strategic or confusion rather than clarity results in a campaign.
Politics is not simply about campaigns, however. It is about a full and rich democratic process. The moment in Victory Lab that is most convincing is when it discusses the work of Jeremy Bird, who helped the Obama campaign engage with communities as they congregate: in churches, schools, barber shops, malls and so on. He argues that the utility of this data is in improving the effectiveness of personal engagement:
“There’s a myth that if you’re using really smart data for targeting, you’re therefore marketing and you’re not doing people-focused, relationship-based organizing. It hopefully helps our organizers and volunteers talk to the right people, and gives us a better sense of who we’re talking to when we call them.”
In other words, these techniques can be used as a starting point for a conversation. The problem is that more often than not that won’t be the approach: it will stay at the impersonal, marketing level. Over time its effectiveness will diminish as people become immune to the messages no matter how cleverly they are crafted around the insights of behavioral science. Healthy democratic politics relies on this constant flow of human relations. Clever and periodic marketing is a very poor substitute.
Alan Gerber and Don Green, the Yale University scientists who have claimed this field as their own, conducted a controlled experiment into the various techniques that worked in getting out the vote. Direct mail increased turnout by 0.6%, phone calls showed no influence but a visit increased turnout by 8.7% in comparison with the control group. Healthy democratic politics is arduous, continuous and personal. The science underpins much of what had been forgotten and, since Obama ’08, is being re-learned. If this is done outside of elections too then democracy will be the winner.
Issenberg’s work is surprising, brilliantly written by a master craftsman, and painstakingly researched. There is an uplifting twist in the tale that I’ll leave for the reader. Lyndon B Johnson’s maxim of “do everything and you will win” is still applicable. It now needs refinement on the basis of the “secret science” of winning campaigns. Perhaps the new maxim should be ‘do everything that works and do it well and you will win.’ The Victory Lab is a must for geeks, gurus, and the rest of us trying to make sense of democratic politics in uncertain times.
Anthony Painter is an author and a critic.