I’m in agreement with most everything in Jonathan Klick’s reply. But I wish to emphasize a couple of his best points.
First, Klick revisits the public/private distinction. This is important because the new paternalists rely extensively on private non-coercive examples, from alarm clocks to ergonomically designed iPods to calorie-controlled food portions, to build support for their position. But to my knowledge, no libertarian has ever opposed these types of private “choice architecture” (although they might be inclined to call them “good product design” or “giving consumers what they want” instead of “paternalism”). The same goes for non-profit private options, such as joining Alcoholics Anonymous, advertising your diet to friends, and so forth.
But the situation changes once the state is involved. In the private sector, you have other options. You can adopt many bias-correcting measures, or some, or none at all. You can refuse to patronize establishments that adopt choice architecture you don’t like. You can search for employers who adopt human resources policies you appreciate. But most people realistically have only one government to choose from; if you don’t like the state’s debiasing methods, too bad. They might allow you to opt out of some measures… or they might not, as a careful reading of the new paternalist literature makes clear.
More importantly, keeping paternalism strictly private creates better incentives. If you fail to correct your own biases sufficiently, you bear the consequences. This at least gives you an incentive to get it right (though not everyone will). State actors, on the other hand, do not bear the consequences of getting it right or wrong. They lack both the information and incentives to pick appropriately.
Second, Klick draws attention to another adverse effect of paternalism: encouraging dependency. The more people believe that the state or some other avuncular party will protect them and guide their choices, the less reason they have to invest in good self-management skills. As a result, paternalism can ironically lead to more of the behaviors it was supposed to correct. This, too, can generate a slippery slope: the worse self-management skills become, the more areas of life seem to invite corrective intervention.
Note lastly that when Klick sanguinely observes that if new paternalism turns out to be a bad idea “we can simply reverse course to fix things in 20 years or so,” he is being facetious. (I wasn’t sure at first, so I asked.) State intervention is notoriously difficult to reverse. Prohibition was the happy exception, not the rule.