Why is controlled regulation so prevalent?

A training participant asked me recently: "Why is controlled regulation so popular when it is clear that an autonomy supportive way of working is so much more effective?" I'll try to answer this question below. Before I do, let me give you a brief summary of what controlled regulation and autonomy support mean. Much research into self-determination theory has shown that autonomous motivation has many advantages over controlled motivation. An autonomy-supportive way of parenting, teaching, coaching, or managing works much better than a controlled approach. The figure below (a translation from a picture from my Dutch book Kiezen voor progressie (2016) summarizes the differences between autonomous and controlled motivation and their different effects:

Knowing is doing

Back to the question: if autonomy support works so much better than controlled regulation, how come the latter is still so prevalent? I see two main reasons. One reason is that we often believe in what we are familiar with. In psychology this phenomenon is known as the mere-exposure effect. It means that we have a preference for things with which we have more experience. If, for example, we have been raised ourselves in a controlled way this increases the chance that we will also use such a style of parenting.We might call this first reason: knowing is doing. But this raises a new question: why were we raised in a controlled way in the first place? Why was this way of working ever popular?

We (mistakenly) think we can see it work

This has to do with the limitedness of our own perceptions and intuitions in matters of psychology. Controlled regulation may not work well in many ways but it does appear to work. The methods of controlled regulation, such as threatening, raising your voice, punishing, promising rewards, etc do affect behavior, and often quickly. A teacher who screams at his students will probably manage to silence them. That teacher may conclude that screaming works and is even necessary to silence students when they are noisy.

But appearances can be deceptive. Much more is happening when the students get silent. These other things which happen may be less easy to observe. For example, they may become scared of the teacher or start to dislike the teacher. The emotional state in which they may get is likely to impede their focus on the subject. Fear and anger doe not mix well with focused learning. Also, it is hard to be interested in a subject when you are fearful or angry.

Furthermore, with this approach, students will not internalize the importance of a quiet atmosphere in the classroom. They keep quiet because of the screaming, not because they want to learn and understand that this needs a certain amount of quiet. Finally, the act of screaming is likely to make the students feel that they can't be trusted. By screaming you imply that a quiet request for silence accompanied with the explanation of why that is needed will not work with them.

Not too long ago a teacher who applies autonomy support in her classes told me that a student asked her in surprised tone of voice: "Ma'am, why are you not screaming at us? Everybody is always screams at us."


Controlled regulation appears to work when we are mainly looking at immediate and superficial effects. The negative side effects which happen are harder to see but very important. Through scientific research we can get a clearer picture of the psychological effects of punishment, rewards, screaming, etc. And that picture doesn't look good.


  1. Thank you, Coert, very insightful! I am also wondering (after reading the new book on Kahnemann and Tversky's collaboration by Michael Lewis "The undoing project") whether regression to the mean might be an additional explanation for why people think that controlling behavior works to motivate people. Lewis cites a story of Israeli army officers who thought that criticism works better than praise because when they praised, they often found that people's performance the next time around was lower and when they criticized the next time around the performance was higher. Kahnemann seemed to have explained this behavior by the fact that you praise when there is an exceptionally good performance -- which means that the next time around is most probably not as good. Encouraging people to take responsibility for their own learning and self-monitoring what they are doing to get better at something is probably even better (and more efficient for the leader and more fun for the learner and more creative and and and).
    Kind regards, Kirsten

  2. Thanks Kirsten, Yes, regression to the mean may partly explain why we overestimate the usefulness of negative feedback and underestimate the effect of positive feedback. Not however that this distinction negative feedback vs positive feedback does not correspond completely to the distinction between controlled regulation and autonomy suppport. In some cases positive feedback may even be an example of controlled regulation. In addition to this, I do not think that regression to the mean explains well other perceived effects of controlled regulation vs autonomy support. The example you mention is a good example of how skewed our observations and intuitions often are and underline the importance of high quality research.

  3. Also, because of the need of power, because of the ego. It is not the same when the subordinate does what he is doing because he has chosen to do it or because he loves to do it. The boss feels the power and feels himself greater when the subordinate does is because the boss said so or because the boss made him to do it. This way the boss feels she is real boss.
    By the way, the need of power is also basic human need. If you have the power, you live longer, you are more happy, satisfied, and so on. Does SDT study the need of power and different controlling mechanisms?

    1. Hi Anonymous, thanks. SDT has several things to say about the need for power. First, there is no evidence that this is a basis human need in the formal sense. Instead, resea5rch suggests it can be a need substitute when the real basis needs have been insufficiently satisfied. Second, a recent study showed that the desire for power is generally more a reflection of the need for autonomy than of the need to dominate other people (see here: https://twitter.com/CoertVisser/status/832244302726168576)


Post a Comment