On paternalism

Arnold Kling imagines the following conversation between parties with opposing views on the desirability of government intervention in citizens' day-to-day decisions:

Libertarian: if a random stranger came to you and forcibly stopped you from drinking a large soda “for your own good,” would you find that acceptable? I would accept advice from a stranger. I might accept forcible restraint from a friend or someone to whom I had given permission to restrain my impulses (like Odysseus with the Sirens). However, why should I want government officials to interfere with my decisions because of my supposed incompetence?

Conly: Government officials are not random strangers. They are experts. That is why the should be allowed to interfere with your decision.

Libertarian: I listen to experts all the time. But what makes government officials so wonderfully expert that I should be forced to listen to them? When it comes to “staying out of debt” and “saving for the future,” are government officials the experts to whom you would have me defer? Seriously?

We should all be able to agree on a few things.  People do, at least on occasion, act in ways contrary to their own best interests.  Perhaps the most compelling evidence of this is ex post regret: "Gee, I guess I should have saved more for retirement."  At the same time, government intervention is imperfect and costly, even at the best of times.  I think it would be a mistake to view either the individual or the government as the ultimate authority in any comprehensive sense.  The libertarian in the above conversation seems to imply that the government should never intervene in individual decisions, and hints that government officials are thoroughly incompetent.  (Maybe I'm reading too much into Kling's post, but the second statement from the libertarian resonates with those views.)

This is one of many issues that is commonly viewed as black and white: should the government intervene in our lives, or not?  Even if we abandon these extremes, the question is often debated as if it were one-dimensional: should we have more or less government intervention?  Even that question is simplistic.  There is no spectrum of degrees of intervention from which we have to choose; rather, it is a matter of considering specific ways in which intervention may improve welfare.  In doing so, why not just think in terms of costs and benefits?  From this standpoint, it is fairly easy to argue against the ban on large soft drinks.  Costs: implementation, enforcement, outrage from affected firms and many members of the public.  Benefits: certain to be positive for some subset of consumers, but limited by consumers' ability to circumvent the law, or the spirit of the law.  This is one place where consideration of government's imperfection really matters.  Even if we agree that it is the government's role to improve public health, is this the way to do it?  Are there better alternatives?  I'm not suggesting that the answer to the first question is a resounding "no," but rather that there is substantial doubt, and that the government should only restrict individual choice if we can make a more compelling case that doing so will create net benefit (which might be the case for something like saving for retirement).

This is the same way I think about government regulation of externalities.  In theory, any time there is an externality, there is a potential welfare gain through government action.  In practice, government action will be costly in various ways and will not hit the theoretical optimum anyway.  So it is only worthwhile for the government to intervene for relatively large externalities.  Determining costs and benefits is typically complex and often controversial; but shouldn't that be the focus of debate, rather than sweeping generalizations about individual autonomy or government effectiveness?