The Economist has an item on what they call "The New Paternalism". It is a commentary on a number of approaches that put a more friendly face to government interference. The paper (as it calls itself) is, of course, still dismissive - being as they are bastions of a a highly principled liberal tradition.
However some of the suggestions are worthy of greater consideration. The ability for a citizen to volunarily restrain themselves from undertaking certain acts in such a way that cannot be broken is a very good idea for people who have a problem addiction (though possibly not as good as some forms of therapy) and especially for people who might have a psychological problem (how to limit a bipolar personality from accessing their credit card). But it does seem open to abuse if it is a substitute where an individual could be reasonably expected to exercise control.
Of far more interest is the recognition that the way choices are presented to individuals will change outcomes. The example given being that if the individual has to opt out of a pension scheme rather than opt in, then participation rates increase. This is one case where the lessons of behavioural economics interestingly intersect with the concepts of institutional economics - the latter being the proposition that the institutions within which markets (the making of choices) operate matter.
The discussion opens up a whole world of realisations that we should be debating a lot more how Governments act than if they should act. In one area close to my heart, business regulation, it is an area way too infrequently considered - so we get lots of advocates aligned for more or less regulation but blessed few discussing how any regulation should be done.