"The same principle applies to roads, as well. Imagine a world without government-built infrastructure. There is no major paved artery that leads to a Costco in a suburb of San Diego. Costco has the option of building such a road, but they know that should the road be built it would also connect their clients to other businesses — which may or may not be competitors. Not building the road means facing a loss: the opportunity cost of an increase in clientele traffic. So, while Costco may not internalize all the benefits of building the road, they may still end up in a better position building it anyways.
The real world is one of disequilibrium and non-optimality, which is why the “economics 101″ public goods argument is not a very good one. As such, there is less of a reason as ever to expect that the market will under provide such goods as roads. There is also plenty of real world evidence of private roads, including large networks of free roads (agricultural roads, including paved highways, in Spain are privately funded, free, and extensive)."
The underlined portions are why I thought it was important to highlight that "underprovide" and "not provide" are two very different things. It's out there and it's natural for a lot of people to talk like this, Ryan. These things happen on the margin. Take a hundred different Costcos with slightly different local infrastructure endowments and costs structures, and on the margin there will be less investment in private roads if they can't internalize all the benefits.
This is why I hate the phrase "public good". I'll get on Ryan Murphy's good side now by saying that I hate the phrase "public good" because it's such an essentialist way of looking at goods and services. Instead we need to think in terms of profit and loss. What profits can be captured? That's what influences decision making.
Jonathan gave a real world example of economics - chicken at Costco. I thought of one yesterday too.
My neighbor, whose house is pretty run down, recently got a brand new roof and is currently in the process not just of repainting but replacing all the siding. It looks a lot nicer than it used to. This is good for me, of course - it keeps property values up.
Which made me think, if we wanted to internalize all the benefits maybe I should cut him a check. It's an obvious positive externality.
Which suggests something else: curb appeal type improvements are probably underinvested in. And it suggests a project for anyone who wants to build a neighborhood's wealth: subsidize curb appeal improvements. Nothing inside the house at all - just exterior improvements. Presumably the gains from interior improvements can be more easily internalized.
I know of at least one example of exactly this approach, which I've always thought would be neat to write a paper on. Does anyone know the example I'm thinking of?