A second familiar objection to paternalism is based on autonomy: simply put, people's freedom to make choices, even non-optimal ones, should be valued per se. Alternatively, as a a sort of mixed utility and autonomy argument, we should value people's preferences for the freedom to make choices. That is, the freedom to choose should also be protected because people value that freedom itself.
The posting title refers to a line from John Robertson, discussing precommitment strategies (i.e., a form of self-paternalism): he suggests that potential gains from such strategies "should not be shunted aside with an ipse dixit about personal liberty." And empirical research may bring into question this second general objection to intervention as well, in at least two ways (thus preventing an ipse dixit shunt).
First, it seems that in fact, people often prefer not to have to make decisions. We prefer to leave our options open (to "leave doors open")--that is, to retain the option of changing our minds, even when that approach comes at the expense of decision quality. In fact, the simple act of considering alternatives and choosing between options can increase the appeal of the foregone option. We ask others to make decisions for us. And, of course, we procrastinate--and, despite realizing that we procrastinate, and making efforts to correct that tendency, it seems that self-imposed deadlines are often ineffective.
Second, when we do make decisions, both the preference for choice, as well as having more options from which to choose, can lead to less utility and worse decisions. Empirical research shows that although people often want a decision to be reversible--that is, they prefer the possibility of changing their mind--they are often less satisfied with their choice when it is reversible compared to when they make a decision and are "stuck" with it. More broadly, recent findings call into question the traditional assumptions that choice generates utility and psychological well-being. In fact, too much choice can be conflicting, difficult, and may lead to decreased experienced utility. Thus, the assumption that people value making decisions is often mistaken; moreover, even when it is correct it can lead to decreased, not increased, utility. (Indeed, in some instances marketers have picked up on this: Iyengar and Lepper cite an instance of Proctor & Gamble reducing the number of available versions of a shampoo brand from 26 to 15, and then "experiencing a 10% increase in sales." Provocative but not conclusive.)
Thus, empirical findings raise questions about another aspect of the autonomy argument. Again, these questions do not mean the argument is invalid, or that paternalistic policies necessarily trump any claim to individual autonomy. The data simply suggest that the standard objections must meet higher hurdles than have heretofore been set, in order to fully persuade.