The thing about BE…[1]: building a better map

200 years ago, Planet Earth still had new places to discover: all you had to do was go there and record them.

In 1798, the young British explorer and cartographer James Rennell returned from an expedition tracing the course of the mighty river Niger in West Africa.

The fruit of this quest – his map of the area – contained one rare and striking jewel: a new range of mountains, which he named The Mountains of Kong (probably following the example of local peoples he encountered). Just as he and his contemporaries had expected: a mighty, mighty river like the fabled river Niger should rise in a dramatic location in a dramatic landscape.

As a result, just about every map of the region in the next hundred years contained the mountains. After all, an English gentleman’s word….

Again and again, cartographers and encyclopaedists repeated Rennell’s discovery – indeed, as late as the 1920s a well-known encyclopaedia was using Rennell’s map as the basis for its descriptions of West Africa.

Yet only 90 years after Rennell’s triumphant return, a French adventurer and soldier Jean Binger actually bothered to visit the place on the map where the Mountains were supposed to be.

And discovered nothing.

Or rather, discovered that there were no mountains at all. Just a plateau at about 1000 meters.

This is the problem with maps we share with others – they can contain very plausible assumptions but unless you are confident about the factual basis they can easily lead you astray.

What the science does for us

This is what contemporary behavioural and cognitive science has done for us – like Binger, it has challenged the assumptions baked into our metaphorical map of human behaviour by “bothering to go there” (i.e. using experiment and observation that can be examined and repeated).

The prime impact has been on the rational agent models of behaviour derived from economics (and utilitarian philosophy more specifically): these assume that humans are utility-calculating machines, using rationality, logic, facts and considered thinking to maximise the value of the choices we make (e.g. to find the best x for the money). Or at least that they should be, if they do thinking properly.

Instead, we have learned how poor a description this is of human decision-making. We have “lazy minds” as Kahneman and Tversky (the Nobel Laureate’s longstanding research partner). Humans are to thinking – so the analogy goes – as cats are to swimming: we can do it if we really have to but only if….

We are more like Star Trek’s Captain Kirk (intuitive, impulsive and emotional) than his colleague, the half-Vulcan, Mr Spock (“illogical, Captain”).

We tend to use shorthands (heuristics like what comes to mind or what’s familiar) to approximate for thinking because a. it is fast b. it works pretty well and c. the difference between pretty well and excellent rarely matters.

And we are “satisficers” (rather than “maximisers”): good enough is mostly good enough.

As a result a lot of the work in this area ends up being concerned with “how smart people do weird shit” (see subsequent blog post for more detail).

How we are “irrational” (actually, a better term here might be “not very good at being rational”). A great example of this is found in the field of Prospect Theory which concerns itself with the rather strange view we have of our future selves and the future value of things. The old saw of “a bird in the hand” being worth two in the bush is not absolutely right in all circumstances but experimental evidence suggests that we significantly overvalue what we have compared to what we might gain in the future.

How our thinking often gets us into trouble and leads us to make mistakes – oh, if only we were less human and more like calculating machines…if only we weren’t swayed by emotions (don’t make decisions angry or sad), by hunger (don’t go food shopping when you are hungry) and so on.

This is really the scope of what is understood as Behavioural Economics.

And it’s a welcome step on for those of us looking to better describe and more effectively change human behaviour in different contexts.

As David Halpern, head of the UK Government’s erstwhile “Nudge Unit” (the Behavioural Insights Team*) describes here, it’s a much better basis on which to design behaviour change interventions such as (in his case) government policies. So too for marketing and organisational change (as my colleagues and I have been exploring in recent years).

A better map means better description of the behaviour you’re targeting and better chances of changing that behaviour.

So far, so good.

But is it a comprehensive map? 

Does it cover all that we know? Or all we need to know? Does it explain all the phenomena that it needs to, to be practically universal?

Not exactly. It doesn’t explain how things spread. It doesn’t explain how things get popular – how somethings grow fast and burn out and others grow in popularity only slowly. In fact, it doesn’t explain very much at the level of groups or families or populations. And that stuff is pretty important – not just because if there’s one really big insight to be had from the human sciences over the last decade it is this: we are a social creature. Equally, most of the users of any map of human behaviour will want to consider things above the level of the individual (even if marketers persist in talking about “the” consumer”): they want lots of people not just you or me to buy stuff. Similarly, while politicians and policy makers may find it easier to think about individuals and individual cases, they are creating policy for large numbers of people who live out there lives in each others’ company. This is why peer-reviews of the impact of BE-led interventions e.g. in healthcare are at best mixed or dependent on other factors

Even at the level of the individual, BE  doesn’t explain why certain shortcuts seem to be hard-wired (i.e. prevalent across human populations). On this, there’s a pretty good piece on evonomics.com by Jason Collins, an australian economist, which discusses what evolutionary explanations might add to the picture created by BE practitioners.

(BTW I feel uncomfortable with the post hoc propter hoc explanations that bad evolutionary science tends to be guilty of – everything isn’t a costly signal which gives an individual performing it a reproductive advantage and very often, evolution can produce things that are good enough, not necessarily the best or the best fit)

Equally, it’s increasingly clear that many of the experiments which reveal “the weird shit” of human decision making do not create universally consistent results. By this I don’t mean that just that the samples on which the experimental findings are based are WEIRD and not reflective of normal human beings (US undergrad populations being probably the least typical examples of human beings to be found on the planet).

Rather, when you bother to go and do the experiments on other populations, with other cultures, you get very different findings. Prof Joe Henrich of British Columbia has built a career doing just this. For example, when the “ultimatum game” (designed to identify inbuilt notions of fairness) was played in communal and co-operative societies like Pacific Whale Hunters or in African Hunter Gatherer groups the results were both different from each other and importantly different from the well-documented patterns of response seen in WEIRD (Western Educated) samples. Similarly, in another experiment when male subjects were asked to elect different outlines of the female body to reflect what they found attractive, different cultures opted for different shapes than those picked by the young WEIRD subjects.

The point being here both that some of the “facts” on which the BE map are based rest on – let’s just say – “wobbly” evidence. And the value of the interventions at a population level, mixed.

No room for “US”

Equally, and perhaps more importantly, BE type insights have no real place in their model for us-factors. Culture – the shared assumptions we have about the world and how to navigate it is a big omission (though hardly surprising, given that BE is essentially the bastard offspring of Economics and Cognitive Psychology).

Indeed, what matters to BE  is – largely speaking – what goes on between an individual’s ears. The mechanisms and default settings of an individual’s mind. The way an individual perceives and acts in the world. This is the prime point of attack it proffers against rational agent accounts of how people decide.

This is why – for example in the classic Nudge Unit Report on BE insights for policy makers, only 1-2 of the 9 insights are anything to do stuff that other agents apart from the individual described. Or why in this more recent slideshare, there are zero references to social. Or why in this otherwise really excellent compendium of BE terms by Alain Samson, social insights are relegated to after thoughts (and normally as with the Nudge Unit report under the rubric of “social proof” – what counts first and foremost is what goes on between an individual’s ears.

This is the for me the big weakness – and very often the big omission – in the BE canon.

So how might we add these?

*In the NYT piece, it is suggested that the NU was set up in 2010 by the Cameron government. In fact, its founding study (MIndspace) and early development were sponsored by the previous Labour Government.