I read an interesting article criticising the "Libertarian Paternalism" of Richard Thaler and Cass Sunstein, by Will Wilkinson, a research fellow at the Cato Institute. Libertarian Paternalism is a proposed ideological and political method that stems from research in behavioral economics and basically offers ways to correct harmfull irrational behaviors of people without depriving them of a final choice (which is what makes this type of paternalism "libertarian").
One cannot deny, and Wilkinson does not, that the awareness of different sorts of irrationality in typical human behavior and of the impact that specific choice-preserving policies can have on peoples' behavior is a great contribution, by behavioral economics. But I have to agree with Wilkinson that this does not pave the way to a new "third way" between libertarianism and other typically paternalist ideologies. I think that "libertarian paternalism" once approved as a government policy method becomes almost indistinguishable from normal paternalism, of the kind that libertarians generally dislike.
When someone reserves the right to dictate when something should be "opt-out" then the distance from being allowed to "opt-out" to not being allowed is much smaller than you think. Given the expanse of today's governmental regulation over our lives it is, I believe, near certain that 'opting-out' of libertarian-paternalist measures will be like 'opting-out' from spammers selling viagra, just too much trouble and usually bound to get you into more trouble. Of course the spammers that use the 'opt-out' excuse for bombing you with pharmaceutical promotions are not correcting social problems, are they?
Sure, from a libertarian perspective, when the government, sets out to correct social ills it seems better if paternalistic policies are 'opt-out' instead of completely obligatory. But then again, the big question has been bypassed completely: if you believe that government is the right organization to identify and correct 'social problems' then, ideologically speaking, you might just forget about choice altogether. That wouldn't be called 'libertarian' though. So, personally I agree that the social technology behind 'libertarian-paternalism' is very interesting, and worthwhile studying/investigating/developing, ie. behavioral economics, but I disagree that it's use points to a new political paradigm. From a political perspective, libertarian paternalism is paternalism, plain and simple.