This paper studies experimentally how people value freedom of choice. Our experiment consists of facing subjects a series of binary comparisons of opportunity sets. Once the subjects have chosen between sets of alternatives, one of their choices is randomly picked: in a first treatment, they have to choose one item among the selected opportunity set (choosing for myself); in a second treatment, the opportunity set is offered to other subjects (choosing for other). This experimental methodology allows us to assess whether people value wider and more diverse opportunity set as well as whether their choices depend on the preferences over items. The results differ according to the treatment. In choosing for other treatment, subjects tend to value larger opportunity sets, i.e., they are more likely to offer others additional option(s) even if more choice does not allow the other subject to choose an item which would give more satisfaction. To sum up, subjects are prone to offering some autonomy of choice. However we also find that subjects may sometimes be paternalists preventing other from choosing their preferred alternative. To contrary, when subjects choose what they can choose, they are more willing to select smaller opportunity set when they are not likely to choose the additional option. We argue that this result is in line with thinking aversion or decision-making hypothesis according to which subjects postpone a decision. Our results may have interesting implications related to the foundation of public policies as well as our understanding of individual decision-making.