Five mistakes we all* make when talking about voters

Linda is a 24 year-old teacher living in London. She is involved in campaigning against tuition fees, attending marches, signing petitions etc. She has always disliked the Conservatives and says she would never vote for them and she thinks Nick Clegg is basically a Tory. Here are some possible descriptions of how Linda voted yesterday – take a second and try to rank them from most likely to the least likely.

A) Linda voted UKIP

B) Linda voted Lib Dem in these local and European elections but plans to vote Labour in the General Election

C) Linda voted Labour and has joined the Labour Party

D) Linda voted Lib Dem

E) Linda voted Green for the Euro elections and Labour for the local elections

F) Linda spoilt her ballot

Done that? Most likely to least likely? Great.

Leaving Linda aside for a minute, why did the voters vote the way they did? And why did you vote the way you did?

There’s a phenomenon in psychology called the illusion of asymmetric insight. People confidently believe that they understand the reasons why other people do things. At the same time, people are sceptical that others could truly understand them. That’s true of college roommates, pro-choice and pro-life campaigners, even people who’ve only met for a minute. In the next few days you’re likely to hear lots of “People voted UKIP because they feel disenfranchised by the political elite and are scared of immigrants” while they will not say “I voted Labour/Conservative/UKIP/Liberal because of my longstanding prejudices.”

We believe that we see the world how it really is. Often that leads people to overestimate how many others agree with them, called the “false consensus effect”. When people see the world differently, there must be a reason. I like Casablanca and if you don’t like Casablanca, it must be because you’re easily bored by black and white films or were distracted when we watched it.

That implies two things. If you don’t have inherent qualities that prevent you seeing reason (e.g. being easily bored) then you should be very easy to persuade. All I need to do is remove whatever bias was preventing you seeing things the way that I see them.  If you’ve read columns predicting that people will agree with us if we only explain that the election is really about this, or that their frustrations are really caused by that, you’ll know what I mean.

Secondly, the more we disagree, the more your views must be explainable. In one experiment, conservatives and liberals were asked for their views on a contentious issue and to assess about how they came to those views. Then they were asked to estimate what the other side thought and how they came to those opposing views. Every time, people predicted wildly more extreme views on the other side, and said they would have come to these views through prejudice rather than the careful consideration that their own group had used.

Part of this is that we underestimate the extent to which others feel doubtful and ambivalent about an issue in the same way we underestimate how often they look in the mirror or secretly feel competitive towards their peers. Because we don’t see others’ doubts, we assume they don’t have them and that leads to the finding that people overestimate how polarised an issue is, even ones as divisive as abortion or race in the US.

That brings me to Linda. It’s a version of an old experiment done by Daniel Kahneman and Amos Tversky but with the stimulus material adapted for the UK in 2014. Did you say B was more likely than D? Then well done, you’re normal. You’re also wrong. The chances of voting Lib Dem will always be higher than the chances of voting Lib Dem and also having any particular plan for how you vote for 2015. But B is closer to what you’d expect Linda to do and so it seems to make more sense.

Take all of this together and you have quite a weird portrayal of the electorate. It’s full of people who I’ve never actually met: sandal wearing, bearded Lib Dems, pin striped UKIP supporters with union jack underwear and so on (“conjunction fallacy”). It doesn’t have any contradictory characters: Eurosceptic Lib Dems, say, or sandal wearing UKIP-ers. The picture includes lots of sensible people who agree with me (“false consensus”). It also includes some people who would be sensible, if only they could just correct their biases (“naïve realism”). My opponents are extreme because of social and economic factors, combined with their sheer wrong-headedness (“false polarisation”). No one else understands this picture quite so well as me, but then again, no one else takes quite so nuanced and ambivalent a view as I do (“asymmetric insight”).

See if you can spot any or all of those in the pages of political analysis printed over the next week. A gold star to anyone who finds an article with all five in.

*me included


Pronin, E., Gilovich, T., & Ross, L. (2004). Objectivity in the Eye of the Beholder: Divergent Perceptions of Bias in Self Versus Others. Psychological Review, 111(3), 781-799.

Pronin, E., Kruger, J., Savitsky, K., & Ross, L. (2001). You Don’t Know Me, But I Know You: The Illusion of Asymmetric Insight. Journal of Personality and Social Psychology, 81(4), 639-656.

Robinson, R.J., Keltner, D., Ward, A., & Ross, L. (1995). Actual Versus Assumed Differences in Construal: “Naive Realism” in Intergroup Perception and Conflict. Journal of Personality and Social Psychology, 68(3), 404-417.

Ross, L., Greene, D., & House, P. (1977). The “False Consensus Effect”: An Egocentric Bias in Social Perception and Attribution Processes. Journal of Experimental Social Psychology, 13, 279-301.

Tversky, A. & Kahneman, D. (1983). Extensional Versus Intuitive Reasoning:
The Conjunction Fallacy in Probability Judgment. Psychological Review, 90(4), 293-315.


Think of Osorno when you hear applause on Question Time

Suppose I tell you that Osorno is a town in Chile. Am I telling the truth? It’s written there in black and white. Weirdly, that last point matters. Black and white is easy to read. Grey and white is hard to read. And if it’s easy to read, more people will believe it’s true.

For the same reason, stocks on the New York Stock Exchange with easy to pronounce codes (like KAR) tend to outperform ones that are harder to pronounce (like RDO) in their first few days of trading.

People like things that are easy for the brain to process, whether it’s clear writing or an easier word to say. Simple things are consistently rated more truthful, more familiar and they make us happier. People even think towns are further away if their names are written in an annoying font.

If you know that, it’s not a surprise that simple political positions tend to poll well. From capital punishment and immigration to controlling prices and nationalising railways, simple often does surprisingly well. Watch an episode of Question Time and see how the more simple the argument is, the more the speaker comes across as honest and somehow familiar, even if you don’t agree with them.

No one in politics has all the simple arguments to themselves. Business and governments normally have more complicated arguments than oppositions and charities. The left tends to get complicated when it talks about international affairs, redistribution or the causes of crime. The right tends to be complicated in its arguments for markets and profits.

Why do any of us stick to complicated arguments – whether they be Marxist dialectics or liberal economics? One answer is that the complicated answers are often right: the world isn’t flat, no matter how honest-sounding flat earth advocates might be. The effect of simplicity may only be weak when people know a lot about an issue or have other cues to follow.

Another answer is that after you’ve learnt a complicated argument, you remember it, while simple arguments – despite seeming familiar – are very easy to forget. So people who’ve learnt something tricky tend to stick with it, while people who’ve agreed with a simple argument might find themselves nodding along to another simple argument a week later.


Alter, A. L. & Oppenheimer, D. M. (2006). Predicting short-term stock fluctuations by using processing fluency. Proceedings of the National Academy of Sciences of the USA, 103(24), 9369-9372.

Alter, A. L. & Oppenheimer, D. M. (2008). Effects of Fluency on Psychological Distance and Mental Construal (or Why New York Is a Large City, but New York Is a Civilized Jungle). Psychological Science, 19(2). 161-167.

Castel, A.D., McCabe, D.P. & Roediger, H.L. III. (2007). Illusions of competence and overestimation of associative memory for identical items: Evidence from judgments of learning. Psychonomic Bulletin & Review, 14(1). 107-111.

Diemand-Yauman, C., Oppenheimer, D.M. & Vaughan, E.B. (2010). Fortune favors the Bold (and the Italicized): Effects of disfluency on educational outcomes. Cognition, doi:10.1016/j.cognition.2010.09.012 Cognition.

Reber, R. & Schwarz, N. (1999). Effects of Perceptual Fluency on Judgments of Truth. Consciousness and Cognition, 8, 338-342

Winkielman, P. & Cacioppo, J. T. (2001) Mind at Ease Puts a Smile on the Face: Psychophysiological Evidence That Processing Facilitation Elicits Positive Affect. Journal of Personality and Social Psychology, 81(6), 989-1,000.


Back again

It has been a little while. I slightly overestimated how much time I’d have for writing while doing an MSc and trying to earn beer money as a freelance consultant.

Now the exams are done, I no longer have to head up to Coventry every other day and there’s just a research project to do. For reasons that are definitely nothing to do with procrastination, I’m going to start blogging a bit more frequently again.

As it gets closer to an election, it gets harder to write anything that isn’t either deeply boring or deeply unhelpful to your own side. So I thought I’d do a few short blogs that are just about applying some of the findings from behavioural and economic science to understanding British politics today. If that takes your fancy, watch this space.


Sources close to the goose

There’s been a longer than usual gap between this blog and the last one and the reason is that I’ve been studying lots and lots of game theory. But during that time I’ve occasionally looked up from the equations and watched a little politics. There have been a few internal Labour rumblings, most notably a “Labour source” criticising Ed Balls’ most recent parliamentary performance before the debate had even finished.

Game theory is a branch of economics that tries to precisely state and predict how people will act when their future is partly determined by what they do, and partly by what someone else does. Game theory has been applied situations from nuclear war to used cars but this incident made me think of a new one: the relationship between party spokespeople and political journalists.

If you are the only person representing a politician, you have a choice. If you act like an automaton, the journalists are likely to be less keen on you and less keen on your boss. If you become friends with them, have a pint and a mild gossip, things might work more smoothly. But if your employer reads “sources close to Bob McBackbench said ‘he never buys his round’”… well it won’t take long before you are tracked down. Whatever happens, you find the balance and you live with the consequences.

When people see critical “sources” in the media, they often assume it is because of some internal factionalism. The interesting thing that you get from game theory is that there doesn’t have to be factionalism: everyone can believe in the leader, everyone can believe that they will be personally better off if only the signed-off lines are spoken and still you can have newspapers full of secrets and stories that only undermine your party.

Take the situation where it’s a two-player game. Suppose the boss is now the party leader and they have two people talking to the press on their behalf: a Lord and an MP. Now these two really do believe in the Leader: in this example, everyone belongs to the same faction. If the party gets a good press and wins the election, all they have to do is divide up Downing Street between the two of them. While that sometimes means giving a favoured journalists a carefully selected secret, most of the time it means sticking to the agreed “line to take”.

So why would either of our players gossip? Firstly, it’s just an easier life. There is no easier job in the world than being the spokesperson that doles out party secrets. By contrast, asserting, hour after hour, that Ed Balls’ speech in the Autumn Statement made you certain of Labour’s imminent electoral victory becomes quite tiring if the journalists aren’t buying it.

Secondly, there might be some personal advantage in a little briefing. My imperfect rule is always that the more and better coverage a campaign guru has, the less committed they are to the campaign they are working on. After all, if you think you’re headed to Downing Street, why would you go out of your way to ensure there’s a flattering profile piece in a Sunday newspaper? But if things are a bit ropey then perhaps you would like to see your name listed as the one the party leader most relies on. The fastest way to get that kind of coverage is to give a few insider details or colourful source quotes to a newspaper or blog.

Finally, and this is how you might justify it to yourself, those indiscretions build up relationships with the journalists that make you more effective. If, in a month’s time, Sophy Ridge is assessing an Ed Balls speech and her source says “actually, this time it was quite good” she will weigh that as far more meaningful than if her source had stuck to the line all along. This makes the party ever more dependent on the person who briefs – and makes the “line to take” ever more irrelevant.

One of the classic findings of game theory is called the “prisoner’s dilemma”: the MP and the Lord would be both better off cooperating but they both know that if one of them sticks to the line, the other would rather be gossiping. If they assume the other one will act on that incentive then they would be a fool not to do so as well: source quotes for the goose can be source quotes for the gander.  In the end, both Lord and MP end up revealing far more than either of them wanted to.

Another more complicated finding is this: suppose the Lord and the MP manage to work together. They realise that getting a big profile in the the lifestyle section isn’t worth it, because while they would enjoy one hit of good media coverage, after that they wouldn’t cooperate and this would hurt their chances of election. The system is working. But then add another adviser to the mix who is also allowed to talk to the press. Now the gains of cooperation are going to be spread thinner: only one of them can be the leader’s main press adviser once they win the election.

In the end, if you rearrange the equations a dozen times, the finding is this: as you add more and more people who can talk to the press, the chances of someone gossiping increase. From that, we can say other things, at least in theory. More spin doctors do not mean better media coverage. A press adviser can be lionised in the media even when the campaign they are working on is headed for disaster. And even a united party can produce pages and pages of briefing against itself.



I’ve become a contributing editor over at Progress magazine – here’s my latest piece for them, on Labour and Tory populism.  It was written before the Autumn Statement – Osborne has been slightly cleverer in his communications than I predicted.




Pincer movement

Two stories caught my eye this week. A Conservative and a Labour MP both attacked Google for having an insufficiently large poppy on the website for Armistice Day. And Ed Miliband criticised short-term loans company Wonga for “targeting children” in their advertising and got a “we’re already on top of this”-type response from the Government.

On the rights and wrongs of the attacks, I think the attack on Google was pretty spurious and this particular attack on Wonga was, if not spurious, possibly a little speculative. But leave all that aside, what does this say about attitudes to business?

Normally, when someone criticises a business, it’s part of a bigger debate. So, if a leftwinger pops up to attack bankers’ bonuses, a rightwinger will then be put up to point to the contribution, not least in taxes, that the City makes to the economy. A local campaigner denounces the company that’s building a windfarm on his or her patch? Very soon a green will push back, if not specifically in defence of the company, certainly in defence of the business model.

That doesn’t happen with Google and Wonga. Crudely, Google is criticised for not censoring more by the traditional right and for not paying more tax by the traditional left. Equally crudely, Wonga is attacked for profiting from poverty by the liberals, and for committing the old sin of usury by the traditionalists.

All or none of these criticisms may be valid and I’ve deliberately stated them in oversimplified terms. But almost everyone has a reason to think that these firms should be criticised. Because people think these firms should be criticised for reasons we believe in, we don’t worry too much if they are criticised for reasons we don’t believe in.

When there’s no debate, really shoddy arguments get through. The “size of poppy = respect for war dead” equation is pretty easy to refute. But the Telegraph and two Members of Parliament from different parties weren’t held back by that: if no one is going to argue back then why worry?*

Most people will say, sod it, both firms are big and ugly enough to look after themselves. I’m a bit worried by the idea of anyone having a carte blanche to politically attack others even if those being attacked are a bit grim (see this from Adam Lent), but maybe I’m missing the bigger picture.

Either way, we’re sending out a strange message.  We’re not saying that companies need to avoid doing things that lead to heavy criticism.  We’re saying that they can be as bad as they want – so long as they keep either the left or the right onside.  Only once they alienate both sides, is it open season.


*With a few exceptions, led by the cursed-to-be-rational @PrimlyStable – go follow her. 


“Some gear shifting in the boldness department”

In the Labour Party, we like boldness. A bit more boldness is always in order. Mark Ferguson has written a call for boldness on LabourList today that’s a good example of the genre.

Mark’s criticism of Ed, and I think he makes it well, is that the boldness of the policies and the boldness of the rhetoric do not match. His answer is to make the policies bolder. The alternative – making the rhetoric match the policies – isn’t considered. It almost never is. Boldness – in our arguments, our policies or our criticisms – is a Good Thing.

By happy coincidence, being bold feels great. If my policies are small and fiddling then I must be small and fiddling. But if my words and my actions are bold then I am not small. Indeed, something as staid as an interest in social democratic politics becomes rather dashing, if you apply enough boldness.

Calling on people to be bolder has a few nice ironies. It’s one of the least risky things any commentator can recommend – who is going to boo me and demand more pusillanimity? It’s also a bit of a contradiction: I want Ed Miliband to believe in himself more, but I’m also telling him to do as I say.

Another question is whether more boldness is more convincing to voters who are making up their mind about Labour. Suppose I tell you you’ll be £100 better off if I win the next election. If you’re convinced, great. But maybe you’re a bit sceptical. Would you be reassured if I came back the next day and said it’s now, £1,000?  How about £10,000?

Being bold feels good when you’re the one being bold. But voters are choosing someone to do something on their behalf. I can’t fly a plane but I imagine pulling off a loop-the-loop feels pretty damn cool. Nevertheless, if the captain starts talking about how bold he or she’s feeling before take off, I think I might want to get off the plane.


Spin cycle

Special advisers have been in the news this week: the government list has gone up, while on the Labour side, Atul Hatwal of LabourUncut has criticised the salaries of some Labour employees (both government and opposition spads are paid for by taxpayers but government spads are employed by the state, while Labour spads are employed by the Labour party). You’ve just missed your chance to apply to be spads to the new shadow transport and development secretaries, or to the new shadow chief secretary.

However, the story people won’t write – and certainly no spad would ever brief! – is that spad-power is currently declining. Parliaments have a cycle. At the start, there’s legislation to pass and reputations to establish. The House of Commons matters. Interesting things are going on in all the departmental briefs, or if they aren’t, then today’s secretary of state for paperclips is interesting because they may be tomorrow’s chancellor. By the time an election is called, the Commons isn’t doing anything interesting, most departments are irrelevant to the election campaign and the spad is just the person that party headquarters phone up when they want a minister to visit a Sure Start centre in Milton Keynes.

The closer you get to the election, the more power shifts to the centre: the leader’s office, the chancellor or the shadow chancellor’s office and the central party officials. That transition is hard but, at least for Labour, it can’t come early enough. Those adverts above should be the last spads we hire and the sooner Ed Miliband, Ed Balls and their advisers transfer over to party headquarters, the better.

Election campaigns work – or don’t – based on trust. You can only work quickly if you trust each other and don’t waste time looking over each other’s shoulders. But in politics it’s natural to worry about each new person you work with. Will they leak things to the media? Can they actually do their job without buggering it up? That might sound like a silly question, but a lot of senior people in politics are appointed because of their political affiliation rather than because they have proved themselves in a similar role. That distrust dissipates over time (or you work out who not to work with).

Last time around we were in office and we knew we were going to lose, so there was always a good reason for the top politicians to keep putting off the inevitable shift to full-time election campaigning. The power will inevitably swing back to the spads after 2015: either for ministers or shadow ministers. Centralising Labour’s political operation as early as possible could help ensure it is the former and not the latter.


Four political thoughts for GDP Day

It’s GDP Day this week, Britain’s three-monthly festival of bad economics. Despite the fact that the estimate will be based on only 45 per cent of the data, some people will make very confident predictions of its political and economic consequences. This time the expectation appears to be that the numbers will be good. The implication is that this will benefit the Conservatives. Good economy = more happy people = more votes for the incumbents.

Four things that make me query this (and they are similar to the reasons why it was right to be sceptical that bad economic numbers earlier in the parliament would hand victory to Labour):

1. We’re now pretty sure that people are “loss averse”: losing a dollar makes someone sadder than gaining a dollar makes them happy. Suppose there’s no growth but a new government comes in and moves some money from project x to project y. It might be the right thing to do, but it will probably piss more people off than it pleases. Without knowing anything else, that government looks like it’s going to lose the next election. Redistributing from very rich to very poor might have a different effect as the first won’t notice it and the second will notice it a lot: but then very rich and very poor seem not to vary their political behaviours very much. If robbing Peter to pay Paul, generally speaking, is a losing plan, GDP growth allows you to pay Paul without robbing Peter. But because Peter gets grumpy faster than Paul gets happy, you need more GDP growth than you think, just to keep everyone roughly as happy as they were at the start.

2. Elections decide the future so some part of the voting decision must be related to predicting which party will make things better in a year or five years’ time. And loss aversion suggests that people will be thinking more about what they stand to lose and less about what they are likely to gain. One of the strange disconnects of political communications is that the speakers always want to talk about the best thing that could happen but the voters talk in terms of the least worst option. Either way, you can’t know how much the Tories might make you worse off without knowing about Labour: it’s the relative number that matters. So a government could be delivering the best economic growth in history; it wouldn’t do them any political good unless people thought the opposition couldn’t manage to do the same. Of course, how a party did before might be a good guide to how it will do in the future. But is a track record of low growth worse than no track record at all? Possibly not, if you’re mostly interested in avoiding a big loss.

3. The scale of the UK’s public sector debt has always suggested that the next parliament will, like this one, be largely about spending cuts and tax rises. The better the GDP numbers, the less this is true, meaning each party will be able to promise a few little fiscal handouts ahead of the election. So far, the Coalition Government has been very bad at maximising the political benefit of its handouts: its increases in the personal tax allowance have gone largely unnoticed while the giveaway over the 50p rate actually managed to hinder the Conservatives’ chances of re-election. If the Conservatives want to say Labour’s tax and spending plans are risky, this is an easier case to make when the cupboard is bare than when the country appears to have a cushion against a shortfall.

4. The Conservative Party, at least sections of it, seem very keen to talk about the degree to which economic growth today proves they were right in the past. Whether this is correct or not, I suspect that talking about yourself and praising yourself tends in politics, just as in life, not to charm your audience.


Ban the (tax) bomb(shell)

One of the pictures above is from the Labour Party speakers’ handbook of 1949. It’s a small book that could be carried around in a Labour candidate’s jacket pocket and contains a little guide to the party’s policies on familiar things like “The cost of living and “Social security” and slightly less familiar policy topics like “Iron and Steel” and “Planning the Private Sector”. This is from the section called “The Conservatives”:

The Tory policy statement…. cynically offers bribes to every sectional interest in the community – regardless of expense. The Conservatives are going to cut taxation [this is followed by a quote from the Conservative policy statement]… then the Conservatives start systematically to bribe individual sections of the community. And all these bribes are going to increase Government expenditure. [Another set of quotes follow]. This programme of bribes is both malicious and dishonest. It is hard to assesses just how much all these things will cost. They must amount to at least £50 million a year. They could only be paid for through an increase in taxation.

The other pictures are the covers of similar attacks made by the governing party on the opposition in the 1992 and 2010 elections. I suspect there hasn’t been a General Election since my 1949 book was printed where the government has failed  to accuse the opposition of making dangerous promises it couldn’t keep (the contradictions on that last point used to be a bit of an obsession of mine).  This is from a mid-noughties briefing for MPs that I wrote when I was working in the successor team to those 1949 Labour researchers – ignore the breathlessly long sentence and note the slight – ahem – surface similarity between our attack on the Tories and events later in the parliament:

The Tories are repeating the same mistakes as 1992 when they promised more spending and less tax, leading to higher interest rates, three million unemployed, negative equity, rising inflation and Black Wednesday – all when David Cameron worked as an economic adviser to the then Chancellor Norman Lamont.

This long tradition could be coming to an end.  This week, Ed Balls proposed changing the legislation that governs the Office for Budget Responsibility (OBR) and empowering it to work out the cost of any pledge made by any major political party in an election. I know, I know, this sounds like a dull little change – but then fixed term parliaments sounded like a dull little change and they have radically changed the way that politics works. This could change how elections work.

Firstly, it would force political parties to say what they intend to do. If it involves money, it has to go to the OBR. If we didn’t put it to the OBR, we were fibbing. No more standing up on the soapbox and saying “This is a dreadful problem” and hoping that everyone listening hears “This is a dreadful problem and I intend to do something about it”. No more “as resources allow” promises. And, if you do work with the OBR beforehand to make sure all the sums add up before you publish them, no more “tax bombshells” either.

Secondly, although the Balls amendment seems ambivalent on this, it would inevitably give the OBR the ability to come up with a forecast of what will happen to the economy differently if a Labour or Conservative government wins office in the future. Most of the time the answer, if you really look at the numbers in detail, will be “not a lot, at least in the short term”.  But oppositions are frequently defeated because their own desire  to sound radical and important conspires with the government’s desire to make them sound dangerous.  An OBR ruling saying, “well this is slightly more pleasant in the short term but has a slightly higher price in the long term” (or vice versa), could change all that.  This time round that would help Labour, but last time it would have helped the Tories.

It would, of course, be bad news for anyone who thinks that large tax cuts or massive public spending programmes pay for themselves in the long run – but if you do think that, you’re going to end up trying to abolish the OBR anyway, so you might as well have the fight now.  The OBR might be completely wrong of course, but comparing two different policy programmes is a much easier task than its “day job” of predicting the future path of the economy.

We spent months of the last election – and it felt longer – debating whether David Cameron had promised marriage tax breaks, what they would really cost and who would pay for them. We still don’t know the answer.  But the debate on whether they were a good idea became irrelevant.  One way to improve politics is to take away all the little proxy issues that exist, both between and within the parties, to keep score.  If we don’t do this,we may just stop caring whether or not a party’s programme adds up.  Of course, there would be many more proxy issues to jump to, but we’d be a little bit closer to having a political debate that resembled the decisions that actually face us.