A punchy, point-by-point guide to anarcho-capitalism. This was recommended to me alongside Structure of Liberty, and I do wish I had read this one first, as it is far more accessible. Friedman writes with the same punchy, accessible style as his father, though perhaps with a bit more of an edge to it. Most chapters are just two or three pages long. Point by point he demolishes the widespread belief that strong centralized government is more effective or beneficial to a modern society than the alternative.

Some points made in the book:

* Anarchy is not chaos. People's tendency to equate lack of central government with chaos stems from the belief that order flows from government. In fact, human societies are self-organizing; historically, governments tend to show fairly late in the game.

* Police and the judicial system are extremely strong candidates for privatization. Barnett spent quite a lot of time on this point as well; it's counterintuitive, but extremely correct. Of course once you realize that it's already the case that the majority of protection services are provided by private agencies in the US, and that private arbitration firms are gaining popularity rapidly, and that most legislated law has its roots in commonlaw, this idea does not seem quite so radical.

* The chapter on national defense is the longest and also offers the least pat answers. Friedman does not draw a strong conclusion that this is something that can be provided without a central government. He does, however, show some options for how it might be possible. He makes a comparison to tipping. Tipping is not required by (legislated) law, nor does it provide any benefits to most individuals (the majority of people who visit a particular restaurant and a particular server will never eat there again). Yet Americans spend billions on tipping each year. If social pressure and a sense of what's right can convince people to contribute to a social good that can easily suffer from the free rider problem, then it seems like it may not be impossible to solve the public good problem of defense via non-coercive means.

After the main part of the book ends, Friedman has an addendum targeted at other libertarians that contains some very interesting material. For one, he argues that the Libertarian party itself is somewhat silly in principle (a centralized, political entity trying to get rid of centralized, political entities) but is an effective way of spreading libertarian ideas. For another, (in the section titled "Problems") he pokes holes in libertarian viewpoints (including his own). He concludes that libertarianism derived from natural rights (the type used by Barnett and many others) is a non-starter, which I found quite interesting, being more or less a natural-rights type myself.