Taking BS to mean behavioral science, the answer to that question is, not quite. But most things are. Most things involve a heavy dose of behavioral science. I would argue, however, the contrary does not apply. I don’t want us to start thinking that BS is everything. It is necessary but not sufficient in many, many cases.
I’d add that BS (behavioral science) without creativity—indeed BS without a tiny little whiff of BS (meaning bullshit)—may be actually suboptimal. If you don’t use behavioral science to expand the potential solution space to a problem by adding a psychological dimension in addition to the other aspects or metrics you’re considering, you’re probably missing a huge opportunity.
This is the vital thing: to a great extent, I think everything is BS. There are huge numbers of problems that persist in the world that probably could be solved much more quickly if people would consider a behavioral or psychological dimension. At the same time, I don’t want us to make the opposite mistake—to immediately go in, look at a problem, and assume that it has to be solved exclusively by the application of behavioral science. A lot of things are a mixture.
If you look at medicine, one of the slightly strange things about it is that they subtract the placebo effect. Now, given that the placebo effect can contribute to a cure, or to the efficacy of a treatment, you’d think people would be trying to actually maximize the placebo effect. What medicine does is say, “We subtract the psychological component of the efficacy of a treatment from the overall efficacy and what remains is the science.” Now, that seems a dubious thing to do. It seems to be self-evident that there are lots of cases where a beneficial combination of a psychological placebo effect and medicinal treatment would be the best solution of all.
One of the great quotes I always use—which, when I first saw it on a PowerPoint slide, I thought was a bit banal, but I now realize is incredibly important—is Harry Truman’s quote, “It is amazing what you can accomplish if you do not care who gets the credit.” Quite a lot of science is so desperate to prove the efficacy of what it itself does—the drugs company can’t claim credit for the placebo effect, they can only claim credit for the medicinal component of the treatment—that we become fixated on proving what we can do alone at the expense of proving and optimizing what we can do together.
This is the really important point about creativity: there are far more good ideas out there we can post-rationalize than there are good ideas we can pre-rationalize.
Let’s take something really simple: a sale in a shop. Now, if you’re a mainstream economist you would look at a sale and you would say, “Well, this is perfectly clear and consistent with mainstream economic theory; you reduce the prices of things and demand goes up.” To some degree, that’s true. I think if you held a sale and you didn’t drop the prices, you could do all the other stuff, and I think people would be fairly pissed off.
But equally, I don’t think it’s fair to say that a sale only works through effectively tweaking the price-demand curve. There are lots of things going on with a sale—there’s scarcity value, the fear that other people might buy the shit that you want, social proof in that there are huge queues of people outside the shop waiting to get in, and an additional element of scarcity in that most sales, although there are exceptions to this, last for a finite length of time. There is a shop on Oxford Street that’s been holding a closing down sale uninterruptedly for about the last five years, but that caters to tourists who aren’t around to notice the inconsistency.
So the most important thing you can do with a sale is not purely the economic bit, and it’s not purely the behavioral bit. It’s both. And in medicine, the best thing you can do is probably combine the drug’s psychological and pharmacological sides.
We need to start making friends with lots of other people who understand that solving problems is more complex and leaves more scope for creativity than the standard models currently allow.
Let me show you what I mean by this. If I gave you and a team of people two candles and I said, “You have to boil this quantity of water using only the candles and a box of matches,” people would probably struggle. And then some shrewd people will come in and say, “Well, actually, this is impossible, because look, the calorific value of the candles is insufficient to heat this volume of water to a point where it actually boils.” So, you put a big tick in the box saying “cannot be done.” And you move on to the next question.
Of course, it’s not quite true, is it? Because the boiling point of water depends on altitude, you could take it to a very, very high place and the same calorific value might well boil the water. If you actually produced a vacuum around the water, it’d be easier still. Or, as another compromise, you could just move the water to France, where, as everybody knows, water boils at 50 degrees. That’s the only thing that explains the shit quality of the tea, to be honest.
There are lots of ways you can solve the problem by simply recontextualizing it and adding new variables to the mix. We rarely notice these things. A lot of people will say, “Okay, this is impossible,” because they’ve failed to add what you might call the altitude or the air pressure component to the mix of variables needing to be considered.
I think there are whole categories in business like this, where some of the world’s greatest geniuses have solved extraordinary engineering problems but have completely failed to take the psychological component into account.
Think of solar panels. I don’t think anybody can fail to be extraordinarily impressed by the improvements made with solar panels in terms of their efficiency, the lower cost to manufacture them, the lower weight. And improvements continue to be made, beyond the point people thought was feasible a few years ago. Engineers have solved that problem, and, geniuses as they are, they deserve credit for doing that. But businesses have missed what you might call air pressure in the equation. They’ve missed, “How the hell do we sell this to people so they put it on the goddamn roof.” It is assumed, still, that solar panels are sold in one irreversible decision to somebody who fixes it irremovably onto their roof with a one percent chance that everything could go totally shit; either your local electricity provider refuses to pay for the electricity you put in or your roof starts collapsing under the weight or you discover you’ve got some hideous problem with beetles.
Now, what we know about humans is they hate making irreversible, five-figure decisions. They really, really hate making those kinds of decisions. So you’ve solved all of the scientific problems, you’ve got the whole how-to-maximize-the-calorific-value-out-of-the-candle stuff done, but you’ve failed to spot the fact that there’s another problem out there that you still have to solve before you can actually have a meaningful effect on personal energy consumption and on mass energy generation.
Nuclear power, you could argue, is even more extreme in that they completely screwed the pooch by choosing the word nuclear, which was associated with bombs, to describe the form of power generation, which it has very little in common with. If you change the psychological frame in which people have to decide, if you change the context, if you change the story, it’s exactly the same as changing air pressure—you change what happens.
Obsessing about the extent to which something you try in the future must make sense in terms of data you have from the past is a massive constraint on innovation and experimentation.
The other point I make is when you’re dealing with the rules of physics, they’re written somewhere upstairs. You can’t change them. Data from the past is necessary and sufficient to explaining what the laws might be in the future, because the laws of physics don’t change (at least at the scale we experience the world). In contrast, the laws of human behavior are highly context dependent, and they change according to things as strange as fashion, whim, storytelling, and mood.
Obsessing about the extent to which something you try in the future must make sense in terms of data you have from the past is a massive constraint on innovation and experimentation. In physics, yes, if it didn’t work in the past, it probably won’t work in the future, and the laws tell you what is impossible.
In anything involving psychology, and, you might argue, certain things involving complex systems, actually trying the same thing again and again and again with the expectation that it might work one time is not necessarily a definition of insanity; it might be a definition of complexity.
Once you accept the fact that we can’t necessarily focus all our time on how to steer the ship in situations where we can also change the weather, or in which the weather can change in ways we don’t expect, we move away from this kind of deterministic, rationalistic obsession with things that make sense in advance.
That’s one of my final creative lessons for behavioral science. Don’t just test the things that make sense. Test the things that don’t make any sense. Then, if you find that they work, you’ve learned something valuable. Actually, you’ve learned something mega valuable, because it’s something that nobody else knows, because the odds are nobody else has been wacko enough to test it.
We did a charity mailing where we tested seven different variables. The one rational one that made sense to everybody was a goddamn disaster. It actually reduced donations by about 30 to 40 percent. All the ones that were wacko improved the level of donation. This is the really important point about creativity: there are far more good ideas out there we can post-rationalize than there are good ideas we can pre-rationalize. The subset of things that we can try that haven’t been tried before that makes sense in advance is very, very small compared to the subset of things that might work in different circumstances.
Don’t just test the things that make sense. Test the things that don’t make any sense. Then, if you find that they work, you’ve learned something valuable . . . because it’s something that nobody else knows.
The bizarre half-sister of creativity is testing and rigorous measurement. They seem like the most different things in the world but they’re actually interdependent. Let me explain how things actually make no sense at all in until you try them.
If I suggest to you that you should get two dishwashers, nearly everybody, either who doesn’t have two dishwashers or who doesn’t have an evangelist friend with two dishwashers, is going to look at me as if I am totally barking insane. But here’s the weird thing: if you have two dishwashers, you never need to unload the dishwasher, and you don’t actually lose any storage space. Because you have a dirty dishwasher, which is where the dirty stuff goes, and you retrieve stuff from the clean dishwasher, eat off it, put it in the dirty dishwasher. When the dirty dishwasher is full, you turn it on, you put the post-it-note that says “clean” on that dishwasher, and then you simply reverse the process. There’s no unloading necessary. There’s no storage lost.
What’s weird is that this occurs to absolutely nobody in advance. It is absolutely obvious in retrospect, but it’s completely nonobvious in advance. And this is why we need creativity and what I call the benign bullshit part of bullshit in behavioral science. Stop obsessing about temperature, the volume of water, and the calorific power of the candles—let’s look for some other variables we might want to introduce.
I didn’t learn any of this at university, I learned it all at Ogilvy. What university teaches you to do is reduce every problem into a two-body problem, which can be solved mathematically to a single, optimal, right answer. Pretend that the problem is that simple, solve for the pretend simplistic model, and then pat yourself on the back. It was only at Ogilvy that I learned, No, the trick here is to go and find some completely different variable that no one’s looking at and try messing around with that instead.
I had a wonderful experience when Daniel Kahneman emailed me and about six or seven other people with exactly the kind of problem we ought to be looking at it. Someone had told Daniel, “A friend of mine sells fruit and vegetables at a farmers market somewhere in Austin, Texas, and he makes pretty large margins”—because, I guess, he’s selling to hipsters, where the lumpier and more deformed the vegetables are, the more they’ll pay—“but the problem is that he can never quite sell out.”
As Richard Thaler rightly explained to the group, scarcity value applies with perhaps pallets of strawberries, and it applies with electrical goods, which are all identical. But with single vegetables and fruit, we tend to assume the last cucumber is probably a bit manky and unwanted, and no one wants to buy it.
Here you had a bunch of people all answering this fundamentally fascinating question about how to sell your last bit of fruit. There were various suggestions, all fascinating. At the very end, I realized that what we were all doing was the same thing that economists would have done. Economists would have said, “It’s very simple, you drop the price.” If you’re an economist, you’re looking at a particular set of variables, which are high-status variables. We were all looking at it through the lens of scarcity value.
Maybe it’s something completely different. Maybe it’s the end of the day and the bloke towards the end of the day is just losing interest a bit and he just doesn’t look that interested in his customers and he starts packing things into the van and then the table is now half covered, so it looks like you’re not supposed to be buying from him. It’s one of the things we have to explore. And it has to be tested as well.
When I wrote my book, I said that if you run a coffee shop, leave the chairs and tables outside even if it’s raining, because from 300 yards away the fact that you have tables and chairs on the pavement signals that you’re open, it means there’s coffee available, because if you closed you would have put the chairs away.
Someone wrote to me and said they used to use that exact insight in reverse. This person worked for a coffee shop and the last thing they wanted was bastards coming in in the last 15 minutes before closing time and ordering a coffee and then sitting around for 45 minutes. Coffee shop owners all over the world are going to hate me for sharing this—all you’ve got to do is take two chairs, stack them upside down on a table.
No one comes in and no one orders a goddamn thing. Simply having two chairs upside down— you haven’t said closed, you haven’t changed the sign—simply doing that tiny little thing is basically enough to prevent anybody coming into your coffee shop.
Everything is BS, but BS isn’t everything, and actually allow space for the other kind of BS. A little bit of bullshit when you’re dealing with a future you can’t predict or fully understand is highly permissible.
What we have to do, effectively, is an element of decision hygiene. We’ve got a very, very weird decision-making process, where the deterministic and the reductionist comes first and the creativity, if it’s admitted at all, comes last.
Look at how government proceeds. It starts off with legislation, then if legislation fails, it moves to economic incentives. If economic incentives fail, it moves on to persuasion. Well, you don’t have to be a libertarian to say that’s completely the wrong way around.
What we need is a long period of recontextualizing the problem through the lens of complexity, the lens of psychology, the lens of behavioral economics, the lens of economics, and the lens of legislation. Drug times placebo is much better than drug or placebo on its own. How do we multiply these various things to actually solve the problem without artificially narrowing it into terms of your imagined high-status solution and implementing that?
When I first met Daniel, in 2009, he said, “I have no hope that we’ll actually make human behavior more rational. What I do hope is that we get more conversation around how people make decisions.” See, people will start saying, “Well, I think I’m going to go on holiday, but I’m wondering if it’s sunk-cost bias, because I’ve already paid the deposit.” As Daniel said back then, if we can simply change how people gossip, if we can change the way people discuss problems so that opening it up to other fields and other variables is seen as a positive, not as a disruption, then we’ve made 90 percent of the progress we need.
I’ve occasionally sat on a board as a nonexecutive director. Do not put a marketing person on a board. Seriously, marketing people will hate me for this; have a separate board for these discussions. There are 10 people all going, “This product isn’t selling, so we need to drop the price.” I’m sitting at the back and what I want to say is, “Have you thought about making it pink?” I know the second I say this it’ll be like passing wind, it’ll be a social embarrassment, all my status will disappear.
If we can rebalance the status of different modes of problem solving, and we can operate these things in parallel rather than in series, we’ll make progress. Just add a bit of creativity, acknowledge that yes, everything is BS, but BS isn’t everything, and actually allow space for the other kind of BS. A little bit of bullshit when you’re dealing with a future you can’t predict or fully understand is highly permissible.
This essay was adapted from a talk given at Nudgestock 2021 and appeared in Behavioral Scientist’s award-winning print edition, Brain Meets World.
