When Improving an Idea, an Object, or a Situation, Think of Subtracting, not just Adding

This is a new format of blogpost: “Paper of the Month”. I’ll regularly spotlight recent research findings in the social sciences, explaining what the researchers found, how they found it, and what practitioners can learn from it.

What are the results?

Adams, Converse, Hales and Klotz (Nature, 2021) 1 have shown that when people try to improve an object, an idea, or a situation, they tend to search for additive solutions (adding something to what already exists) rather than subtractive solutions (removing something that already exists). This tendency is particularly strong when people have a lot of information to process, and when they do not have the opportunity to experiment with multiple solutions. On the contrary, prompting people to consider subtracting (and not just adding) appears to mitigate the effect.

How did they show that?

Adams and colleagues used multiple laboratory experiments. For example, in one of them, they showed participants a Lego structure, and asked them to change it in such a way that the roof could hold a masonry brick without collapsing over the head of a figurine. For each brick that they added, participants would lose money.

screen reader text
Lego structure

The best solution is to remove the brick circled in red: Participants would not add any brick, and therefore not incur any cost. When participants were not instructed to consider removing bricks, only 41% of participants found this solution: Most added bricks, and therefore lost money. In contrast, when participants were prompted to consider adding or removing bricks, 61% found the optimal solution. In other words, if not explicitly mentioned, most people seem to overlook the benefits of subtracting (rather than adding) elements.

In another study, the authors tested how having the opportunity to experiment with the task would change this tendency. They showed participants a digital grid of white and green tiles. Participants could click on any box to change its color. Participants’ goal was to make the grid symmetrical, both on the left-right axis and the top-bottom axis, using as few clicks as possible. Again, the correct answer was to remove a few green tiles (circled in red), rather than adding more of them.

screen reader text
White and green tiles

When participants were not given the opportunity to practice, only 49% of them found the optimal subtractive solution. When they were given three practice trials however, 63% of them found it.

Finally, Adams and colleagues asked people to solve the same grid task, with a twist: A subset of participants was asked to complete another task while solving the grid. Asking people to do two tasks at the same time is a common way of manipulating cognitive load, and a way to see how people perform when they have a lot on their mind. They found that participants under cognitive load were less likely to identify the subtractive solution: This suggests that “adding” is an easier, simpler solution for people to reach.

What are the implications of these results?

If people tend to add things every time they try to improve something, it poses significant risks to organizations: Projects, designs, and processes might become more and more complex and bloated, ultimately leading to a waste of resources, and increasing inefficiency.

Why is it happening?

Let’s consider a team that has been developing a new product for months. At each round of brainstorming, the temptation might be high to add new features rather than removing some. There are various psychological mechanisms that would explain this tendency.

First, the “endowment effect” 2 and “loss aversion” 3 both describe how people grow attached to things they own, and are averse to letting them go. Removing features on which the team has spent resources (time, efforts, money) might therefore be psychologically costly, which would lead people to add new unnecessary features rather than removing obsolete ones.

Another reason why people would prefer an additive solution to a subtractive solution is to avoid conflict. A solution that would remove something that was previously suggested and implemented by one(s) of your colleagues risk being misinterpreted: The people at the origin of those features might feel slighted, or view the decision as an attack on their competence. In this case, an additive solution might be perceived as preferable because “easier” and “more peaceful”, even if it is less efficient.

Finally, people might be biased towards what is already there. When people do not know why things are the way they are, and how they interact with each other, they might be tempted to take what is there for granted (“it’s here for a reason”). In this context, they might view subtractive solutions as risky, and additive solutions as safer (“removing this thing might break everything, so let’s add something new instead”).

How can organizations avoid this tendency?

First, creativity is not only about creating new things: It is also about identifying inefficiencies and redundancies and learning how to accomplish the same thing with fewer elements. This form of creativity needs to be acknowledged and rewarded if organizations want to avoid the problem of “adding always more”.

Second, people should be mindful of the opportunity costs of adding new features. When people add something while the more optimal solution would be to remove, they incur two costs: The cost of the addition, and the cost of keeping something inefficient in the system.

Third, the paper of Adams and colleagues shows that subtractive solutions are less intuitive: It means that people will need time, mental space, and support to reach them. They first need time to recognize the disadvantages of an additive solution: Here, giving people the opportunity for trial and error might be useful. People will then need the headspace and bandwidth necessary to envision what a subtractive solution would look like: They need a comprehensive picture of the situation, an understanding of the different parts and how they interact together, and of the pros and cons of removing any of them. Finally, people will need support to endorse a subtractive solution: Because people are averse to loss, and because they often feel ownership for what they have created or built, they will often need a bit of time to accept the idea that they should get rid of a feature on which they have spent time and energy.

How can managers facilitate the adoption of subtractive solutions? As often, communication is key. They should first remind team members that removing something doesn’t mean that their work is not appreciated, or that they wasted their time: Just because something is no longer useful doesn’t mean that it never was. They should then emphasize that removing features is not a signal of failure: Organizations grow and learn by letting go of what is unnecessary. Finally, they should reward and promote employees who take the time to critically re-evaluate how things are done, and who are not afraid to suggest subtractive solutions when they are needed.


Adams GS, Converse BA, Hales AH, Klotz LE. People systematically overlook subtractive changes. Nature. 2021 Apr;592(7853):258–61.

Meyvis T, Yoon H. Adding is favoured over subtracting in problem solving. Nature. 2021 Apr;592(7853):189–90.

  1. Materials, data, and codes are in open access on the OSF. ↩︎

  2. Endowment effect: When people own something, they are more likely to retain it (i.e., they will typically ask more money to get rid of it than they will pay to acquire it). ↩︎

  3. Loss aversion: People are more sensitive to losing things than to gaining equivalent things, and tend to make decisions to minimize losses rather than maximize gains. ↩︎

Found this post insightful? Get email alerts for new posts by subscribing:

Zoé Ziani
Zoé Ziani

PhD in Organizational Behavior