Over at Libertarianism.org, Matt Zwolinski gives six reasons why he does not believe that the non aggression principle can function as the sole principle of libertarian theory. His six reasons are that the principle:
1. Prohibits all pollution
2. Prohibits small harms for large benefits
3. Adopts an all-or-nothing attitude toward risk (either, depending on how one interprets the principle, prohibiting or allowing all actions that impose risk on others)
4. Does not generate a prohibition on fraud
5. Is parasitic on a theory of property (since defining force depends on a prior theory of property rights), and
6. Seems to generate unacceptable positions as to children (such as allowing a parent to starve his child).
I agree that all of these are problems for the non-agression principle. This is one of the reasons that I moved from a natural rights foundation for my libertarianism to an indirect consequentialist approach. In my view, an indirect consequentialist approach addresses each of these points directly and gets at attractive results.
(Indirect consequentialism holds that individual actors should not calculate utilities but instead should follow rules or other guides. While rule utilitarianism is a kind of indirect consequentialism, I advocate a version that derives the indirect aspects not from an assumption that we should follow rules, but based on an argument that following rules leads to the best consequences.)
Of course, many people, upon hearing a reference to consequentialism, immediately claim that it leads to all kinds of unacceptable results. Many of these people are libertarians. Strangely, these are people who believe that liberty produces great results. But these are matters for another post.