Addition bias
You probably know this sight from the airport. You are standing before the check-in, and in front of you a zigzag of belts and posts is laid out, the barriers that guide people in a snaking line before they reach the desk. Except it is the middle of the day, the terminal is nearly empty and there is nobody but you. In a moment like that I walk straight ahead, under the belt, right to the gate. Walking the zigzag when there is no queue would feel a little pointless.
And yet every time I feel a small pang, as if I were breaking some rule. Supposedly nobody is watching, supposedly those posts in the empty terminal serve no one, and still walking straight takes a tiny decision from me. Someone set up those barriers once, because there was a crowd. The crowd is long gone, and the barriers still stand. Nobody takes them down, because taking them down is someone’s decision, while leaving them is nobody’s.
I think this is what most of what builds up in companies and in our everyday work looks like. Some process, an extra meeting in the calendar, a new field in a form, another tool, and on top of that one more layer of approvals. Each of these things was created once for a sensible reason, for its own crowd. The crowd dispersed, the reason expired, and the barrier stayed. And almost no one asks the question that begs to be asked: why, when we want to improve something, do we almost always add instead of subtract?
Because it is not a matter of laziness or bad will. It turns out that adding starts in us much earlier, before we even get to any decision at all. It starts in the mind itself.
The mind that does not see the minus
A few years ago a group of researchers at the University of Virginia decided to test something seemingly trivial. They gave people a Lego structure in which a platform was meant to hold up a heavy brick, but one of the supports was too narrow and the whole thing wobbled. The simplest fix was to remove a single brick so the top settled on the wider column. What is more, every brick added cost the participant money, while removing one cost nothing. With no hint at all, 41% of those tested removed the brick. When the researchers added one sentence, that removing bricks is free, the share who subtracted rose to 61%. The results were published in Nature in 2021.
In another approach they asked people for ideas on how to improve a certain university document. Out of more than six hundred suggestions, only roughly one in ten involved removing something. All the rest was adding. The most important thing, though, is not that people preferred to add. The most important thing is that they did not consider subtraction at all until someone reminded them of it.
Researchers called this the addition bias. When we have to improve something, our mind by default searches for solutions that involve adding something, and subtraction simply comes to mind less often, because it takes more effort. And what matters, the busier the participant’s head, the less likely they were to even think of taking something away. Under pressure and in a hurry, subtraction is the first thing to vanish from the radar.
I catch myself doing this more often than I would like to admit. By nature I like to put the world in order. I take something that barely works and shape it into a process, a structure, some sensible order, and that genuinely draws me in. Except the things I create never feel good enough. I always have the impression that I could add one more step, throw in a safeguard, gather a bit more data before I decide. I had to learn at some point to tell myself that this is good enough and that I am leaving it. Because otherwise improving has no end, and improving almost always first suggests that I add something.
And if this tendency sits in each of us, then in someone who makes decisions for an entire company it grows even stronger.
Why leaders add
There is a study about football penalty kicks that sheds a lot of light on why we prefer to act in the first place. Economists analysed close to 290 of them in the top leagues and worked out what pays off most for the goalkeeper. It turned out that statistically it would be best to stay in the centre of the goal, because that is where the most shots are saved. And yet goalkeepers throw themselves left or right in some 94% of cases. The authors of the study explain it simply. When you let in a goal while standing still, it looks as if you did nothing. When you let it in after a spectacular dive, at least you tried. A goal conceded from inaction hurts more than a goal conceded from action, even though the scoreboard reads exactly the same.
With a leader it is similar, only the stakes are higher. On top of that comes something psychologists call the omission bias. In short, when we actively remove something and it goes wrong, we feel guilty that we are the ones who did it. When we leave things as they are, or add another safeguard, and it also goes wrong, we have the sense that at least it was not our fault. A meta-analysis of more than twenty studies showed that we judge harm caused by active action far more harshly than the same harm that came from passivity. In a leader’s decisions this asymmetry becomes very costly, because removing something is always a form of action, one you can later be held to account for.
Plain relationship arithmetic plays a part too. When you add people and budget, you gain allies, and when you subtract, it is easy to make enemies. And there is one more thing, more human. We simply want to see value in what we do. When I add something, it shows that I worked, that something bigger and more elaborate came into being. When I take something away, from the outside it looks like almost nothing, even though very often that was the harder move.
You see it best not in big strategic decisions, but in something as ordinary as the company calendar.
How meetings multiply
I have the best example from my own company. A few years ago, when we started to grow faster, we were shaping the way we communicate and synchronise with each other. At first one meeting at the start of the week was enough. Then a second one came, at the end, to wrap everything up. Then smaller syncs inside particular areas and subteams. None of those meetings was stupid. Each one on its own made sense. Except at some point I looked at the calendar and saw that those meetings had grown really numerous (more than any of us had consciously planned).
And what did we do first? For a long time we simply added more, because each one solved some real communication problem. Today it works differently with us. We settle most things asynchronously, and in the calendar there is essentially one main slot in which we all meet, plan the week together and sum it up at the end.
Those were our barriers in the empty airport. We set each one up when there was a crowd, that is, a real problem to solve. The crowd dispersed, and they still stood, because removing any single meeting always required someone to consciously say that we no longer need it. And that, as we already know, is harder than adding another one.
This is not only my story, by the way. Bain calculated, on a sample of three hundred large companies, that the average organisation loses around 21% of its productive capacity to its own complexity, that is, to redundant meetings, layers and procedures. That is more than one day a week in which people service the company instead of doing what the company exists for. The authors point out one thing. This organisational complexity does not come from any breakdown or anyone’s stupidity. It is a natural consequence of growth. The bigger the company, the more places where someone once, for a good reason, put up a barrier. Process bloat in a company is almost never anyone’s deliberate choice. It is the sum of many small “let us add this”, each of which on its own was reasonable.
Except those percentages and hours are not an abstraction from a spreadsheet. Behind each of these layers stands a specific person.
Who carries all of it
Every layer that gets added has its owner. A report that someone fills in every week, though they no longer remember who actually reads it. A meeting someone logs into, because disappearing from it would look odd. A field in a form that has to be filled, because once, long ago, someone needed it. All of this adds up to someone’s ordinary morning.
It hits the middle managers hardest, the people who have the most to handle and the least freedom to take anything away. McKinsey asked several hundred middle managers what spoils their work the most. The most common answer, more common than anything else, was bureaucracy: too many meetings, emails and approval processes. 44% of them named it. On top of that comes the daily switching between tools. An analysis of the work of several dozen teams in large companies showed that a person switches between applications and tabs around 1200 times a day and loses, on finding their bearings in all of it, roughly four hours a week.
It is easy to think about this in terms of productivity and lost hours, but for me something else matters more. All these layers wear people down in a way you cannot see from the outside. Nobody rebels against a single field in a form or a single meeting a week. It is just that in the evening a person is a little more tired and a little less sure what they actually did half of those things for. And fatigue like that, spread across a hundred small things, is the hardest to subtract afterwards, because nobody knows where to start.
So if adding costs so much, why is subtracting so hard that it takes a separate piece to write about it? Because taking away is not the same as tidying up.
Subtraction is a decision, not tidying up
Let me show you an example that says a lot about how much this is about mindset rather than the numbers themselves. McKinsey described a team that was given the task of finding savings in a budget. When people looked for what to cut, they found less than 5%. Then the same task was framed differently. Do not cut, but build this budget from zero, item by item, and justify each one that is to stay. The same people then found more than 40% in savings.
It was the same task and the same people. Only the frame changed. When we subtract, we feel that we are losing something, and loss hurts. When we build from zero, we lose nothing, we simply choose consciously what is worth adding back. And suddenly subtraction stops chafing. I came to this through practice rather than from books. It is easy to read in some wise person that less means better. It is harder to believe it until you see in your own company that improvement more often means taking something off than putting something on. Today, when something is not working with us, the first question I ask myself is rather “what can we take away from here”, not “what else can we throw in”.
I just would not want you to walk away from this text with the conviction that the answer is “cut everything”. That would be just as naive as adding endlessly. Let us go back for a moment to the barriers at the airport. They really were set up once for a crowd. If you take them down at rush hour, chaos breaks out. There is an old rule that you should not remove a fence until you understand why someone put it there. In a company it works the same way. Some “redundant” processes are buffers that protect you from a risk invisible day to day, one that can come back with double force. In the theory of reliable systems it has long been known that removing slack and redundancy can make the whole thing fragile.
That is why subtraction tends to be harder than addition. It requires you first to understand why a given thing came into being at all, and then to take on the risk of removing it. Adding itself is not the culprit here. The trouble is that adding is our default setting, while subtraction is something we still have to learn. Maybe that is exactly why simplifying organisations so often ends badly, when someone treats it as an even cut of everything bit by bit, instead of a conscious choice about what stays and why.
And there is one more layer that we add most eagerly and find hardest to give up. Control.
The control we are afraid to give up
My own relationship with control taught me the most about this. I carry a fairly deep rebellion against it, both the kind that comes from outside and the kind a person tries to impose on themselves. If I had to convince myself that I must watch and control myself in some area, it probably would never happen.
I have lived since birth with an illness that meant I did many things not because I wanted to, but because I had to. Out of that necessity, discipline was born over time. And discipline, although it is itself a form of control over oneself, had in the end to give way to something else, to an acceptance that not everything can be held in your grip. This applies to both my professional path and my private one. I carry it straight over into the company. I believe that any form of hard control sooner or later loses, and the only thing you can really build an organisation on is trust.
And here the whole text comes back. Because an extra process, another approval, one more control report, is nothing other than adding control. It is exactly the same tendency to add, only dressed in the word “safety”. When I do not trust, I add a barrier. When I trust, I can take it down. In companies I very often hear “let us not touch this, it has always worked”. I understand the temptation. But if we stop improving even the things that work, including improving them through simplification, it will be hard for us to remain a company that does anything differently and better than the rest.
I have no ready recipe for you here. What I do have is one question that I ask myself more often than is comfortable. What zigzag of barriers stands today in your company in an empty terminal, kept only because there was once a crowd there, and nobody so far has dared to take it down? And next time you are improving something, will the first thing that comes to mind be to add something again, or will you ask what could be taken away?
