Bad seafood, losing gambles and rat brains

Behavioural economists love talking about risk. Politics isn’t very good at talking about risk: governments have to pretend there’s no possibility that their plans will fail, commentators have to assert that – if only they followed my plan – my party would win a landslide. Bad results are seen as proof of bad ideas and bad character – and rarely of bad luck.

I like thinking in terms of risks so here are three things we have learnt from studying risky decisionmaking that might have some application to politics and elections.

1. Going back to a bad restaurant

Suppose there are two similarly priced seafood restaurants in a town you’re visiting for two nights. For each restaurant you’re interested in two facts. The first one is obvious: what sort of quality can you expect as an average customer. The second fact is how much the quality may vary. It’s particularly true of seafood but, generally, people are more worried about bad risks (food poisoning) than they are about good ones (occasionally the normal chef’s incredibly talented cousin drops by and cooks something special).

When you arrive in town, you assume that both restaurants have an equally good average and are equally variable. You pick one at random but you’re disappointed: your fish is overcooked and the staff are a little rude. Now it’s the next night: which restaurant do you go to? You’ve now lowered your expectation of what the average meal is like at the restaurant you visited. But this bad restaurant has become more attractive in a different way: you now have a better idea of the range of possible outcomes that could happen on a return visit.

After four years in office, some people, probably many, will think that the Cameron Government has been worse than they expected. The Tories weren’t didn’t reveal beforehand the extent of the public sector cuts and tax rises that they later thought appropriate. But the fact of being in government has reduced the variability  in the possible outcomes associated with a Tory victory in 2015. This applies to good outcomes and bad outcomes. Opponents would be more worried in 2010 than in 2015 that the Tories would scrap the NHS. Supporters would be more excited in 2010 than in 2015 that the Tories would dramatically reduce taxes. But – as in the case of seafood – bad risks tend to outweigh good ones, perhaps giving Prime Minister David Cameron a small net advantage over David Cameron as Leader of the Opposition.

2. People gamble when things might get worse, not when they have got worse

One of the most important insights of behavioural economics is that people act differently when the choice is getting £1 for certain or a 1% chance of winning £100 than they do when the choice is between giving up £1 for certain or a 1% chance of losing £100. When it’s a choice about gains, people tend to pick the safe option. When it’s a choice about losses, people are more likely to pick the risky option.

What does that mean for politics? If people think the next parliament is about gains then they may be more likely to back low-risk plans or incumbent politicians. That’s true even if they’ve suffered painful losses through this parliament. On the other hand, if people think the next parliament is going to be about either accepting inevitable pain or taking big risks to avoid pain, then they might pick the high risk plans and the insurgent politicians. That’s true even if the last four years have been pretty good for these people.

Now, I’ve warned myself before about dealing in archetypes, but as politics is currently obsessing over small-c conservative, working class voters, these two models made me think of two different caricatures of the same demographic. The caricature of the traditional working class Tory is someone who’s got their own council house and hopes for a pay rise at some point in the future and he or she doesn’t want any gambles.

On the other hand, if the choice is between accepting that your area is going into sustained decline or taking a risky gamble to prevent it, people might well be ready to “set [their] life on any chance, to mend it or be rid of it”. That might be UKIP, it could be an equally immoderate position on the left.

However, just because those two archetypes are equally easy to think up, doesn’t mean there are equal number of voters that get anywhere near the two caricatures.

3. Fears about parties can reappear very quickly

Researchers can teach rats to be scared of a particular sound by repeatedly making the sound and then giving the rats an electric shock. They can then unteach the rats by making the sound without making the shock until the rats start to ignore it. These rats seem to be completely back to how they were before the experiment. But… if the sound and shocks start to coincide again, even just briefly, the rats quickly go back to being scared of the sound.

What does this have to do with human politics? The fears of rats and the fears of people aren’t as different as you might think because our fears are controlled by an old part of the brain – the amygdala – which does the same thing in birds, rats, rabbits, monkeys and humans.

If fears can remain dormant and then be quickly re-awakened, party rebranding will take much longer than it might appear on the surface. In the 2010 election, our internal research showed that some people were still worried about the Conservative Party based on something that happened in 1992: the wave of home repossessions that followed Black Wednesday. For the last few years, there has been no reason to associate repossessions and Conservative-led Governments. But the Tories would be unwise to think that the toxic link between Conservative-governments and home repossessions  has disappeared: if rates rise in the next year, and repossessions rise as a consequence, Osborne could be in for more trouble than he expects.



LeDoux, J. (1998). The Emotional Brain, Simon & Schuster: New York.

Quattrone, G.A. & Tversky, A. Contrasting Rational and Psychological Analyses of Political Choice. American Political Science Review, 82:3, 719-736.

Rust. R., Inman, J.J., Jia, J., & Zahorik, A. (1999). What You Don’t Know About Customer- Perceived Quality: The Role of Customer Expectation Distributions. Marketing Science, 18(1), 77-92.

Tversky, A. & Kahneman, D. (1992). Advances in Prospect Theory: Cumulative Representation of Uncertainty. Journal of Risk and Uncertainty, 5, 297-323.