Tuesday, 6 April 2010

Behavioural politics, day 1 of 30

Today Gordon Brown announced that there will be a general election in the UK on Thursday 6 May.

Over the next 30 days, I'll be analysing at least one policy each day from a cognitive/behavioural point of view. Voting decisions are, in many ways, like purchasing decisions - and many of the same cognitive biases apply to both interactions. In each case there's a seller and a buyer, and the seller's job is to present a package to the buyer that maximises the likelihood of "purchase". I'll be showing how the parties can do that.

So please feel free to treat this as a critique of the UK's party manifestos, or as a 30-part tutorial in how to apply cognitive insights to politics. Perhaps a new discipline will emerge: let's call it behavioural politics.

Today, as no real policies have been announced yet, the focus is on the overall campaign messages. Key events of the day (note that the linked page may update in real time - I will try to find a permalink to the content).
  • Social proof: Cameron's use of "hope, optimism and change" - creating an association with the successful campaign of Barack Obama
  • Loss aversion: gay people noticed the omission of "gay and straight" from Cameron's speech much more than they'd have noticed its inclusion if he had remembered to include it
  • Game theory and multiple equilibria: the Liberal Democrats are keen to ensure people consider them as a real option with a chance of winning, otherwise they will be ignored in a self-fulfilling voter strategy
  • Authority bias: we tend to trust people more if they express high levels of confidence. Thus, Gordon Brown is visiting constituencies he has been given little chance of winning, to send a signal of confidence.
  • Storytelling: David Cameron knows how effective stories are, and has been telling the story of how his family has benefited from the NHS. We are much more likely to believe assertions that are supported by stories, and this helps make Cameron's commitment to the NHS appear more credible.
  • Social proof again: the Conservatives are playing up the fact that three more business leaders have signed up to their commitment to cancel the rise in national insurance.
  • On the same issue, reinforcement bias: twenty CEOs signing a letter on Saturday followed by three more today and another four on Thursday is much more persuasive than simply having all twenty-seven signing up in the first place.
  • Saliency from simplicity, related to the conjunction fallacy: BNP leader Nick Griffin's key policy is to pull British troops out of Afghanistan. By focusing on a single policy, a single-issue party may capture voters from other parties who have multiple policies, even if they overlap with that single issue. That is, people might vote for a party who promises only to leave Afghanistan, in preference to a party who promises to leave Afghanistan and improve the NHS. Of course this is especially important for Nick Griffin as he does not want anyone to notice his party's real policies.
  • Labour has also used saliency by trailing a policy announcement on constitutional reform; which gives voters an anchor and reminds us that there is something more to talk about than just yet more meta-analysis about the campaign itself.
  • Confirmation bias: In most statements, we listen to what confirms our existing beliefs. Therefore, Ed Miliband asserting that Gordon Brown was not intending a dig at Cameron's rich upbringing, confirms that Brown was indeed intending a dig at Cameron's rich upbringing. Note that the Daily Mail, in protesting against Brown's "spitefulness" runs the risk of merely entrenching this assumption about Cameron's background.
  • And there are lots of examples of people trying to defuse each other's future appeals to cognitive bias: Osborne warns that Labour will try to use fear as a campaigning tool, while Ed Miliband calls David Cameron "a salesman". We know that the more awareness we have of our cognitive biases, the less effective they are; though it's unlikely that we'll remember these warnings long enough to make any difference.
So if I were running a party at this election, what would be my behavioural policy of the day? Well, like the real parties, I'd be cautious about being too specific at this point - except with one or two highly uncontroversial policies which might make my offering look more concrete, and therefore more honest, than my competitors.

In choosing my campaign themes, I'd use loss aversion by focusing on a couple of key weaknesses of the opposition, which I can keep reminding people through the right priming and use of language, and pick a couple of my own perceived weaknesses, especially those which are unconsciously believed by voters, which I could expose and defuse by my actions during the campaign.

I'd use endorsements from trusted individuals as social proof - choosing individuals who are trusted by the median (swing) voter, rather than by my own core voters. And I'd tell stories which express my successes (if I'm the incumbent) or the government's failures (if I'm the challenger).

Finally, I'd choose a meta-narrative - at first, more for the benefit of the media than the voters directly - which will be reinforced through continuous contrast effects and confirmation bias throughout the campaign. The government's needs to be trust, safety (contrasted with fear); the Conservatives' needs to be change (contrasted with disgust); the Liberal Democrats' would be credibility and a clean break (contrasted with tiredness and corruption) and the small parties would be a salient single issue (contrasted with outrage at the government).

And let's not forget that recency bias means whatever happens today is of limited importance. Votes will be cast based on unconscious beliefs - which can be influenced to some extent by priming and trust during the campaign - and on whatever concrete policies are overtly discussed in the last few days before polling day.

Ratings so far: Conservatives 7/10, Labour 7/10, Lib Dems 5/10.

1 comment:

Min said...

"Saliency from simplicity, related to the conjunction fallacy: BNP leader Nick Griffin's key policy is to pull British troops out of Afghanistan. By focusing on a single policy, a single-issue party may capture voters from other parties who have multiple policies, even if they overlap with that single issue. That is, people might vote for a party who promises only to leave Afghanistan, in preference to a party who promises to leave Afghanistan and improve the NHS."

There is another aspect here, I think. Politicians love to make promises. Much more than they keep them. In my mind, then, a politician who makes a single promise is more likely to keep it than one who makes the same promise, but also makes others. The single promise is more believable, as is the candidate who makes it.