27 October, 2008

Costs and Benefits

Matt Yglesias and Ezra Klein are shocked and dismayed that Kevin Drum opposes California's High-Speed Rail Initiative. I don't know about the details of this ballot proposition, but Drum's absolutely right that there are good transit projects and bad transit projects. In fact, we know very little about what the good transit projects are. There's hardly any solid empirical work documenting the benefits from public transportation projects. That's not to say that there aren't any--there's a lot of theory to suggest that they are. It's just that we don't know in any rigorous way which ones they are. Government spending is expensive. Every dollar spent by the government is a dollar not spent by a real person. It's easy to get caught up in ideas--transit, health care, climate change--and demand that something, anything be done. But we should always try to take a broader view when evaluating particular policies.


Elliot said...

I agree that the "we must do something; this is something; thus we must do this" mentality is not healthy. (see: offshore drilling)

But I take Yglesias' point to be that while the project is not perfect, its still reasonably good. And at a time when we could really use some heavy duty fiscal stimulus, it makes sense for the government to be directing that money towards useful infrastructure sooner rather than later. And I don't really know either - maybe this is a bridge to nowhere type project. But 3.5 hrs to get from LA to San Fran for an estimated price of 50$ (those are the numbers Klein uses, anyway) seems like a pretty useful project to me.

In a larger sense, there is always a balancing act between more and better information and the need for action. As time approaches infinity, our understanding of the optimal public policy may approach perfection. But operating under time constraints, how long do you put off needed projects in order to accumulate more empirical evidence? And what if that evidence can really only come from trying and failing? I don't mean those as rhetorical questions; the debate going on over prop 1 has a lot of bearing over how we are going to see future government projects such as cap n trade or alternative energy grids that don't necessarily have a lot of precedents.

spencer said...

Right, I don't know the specifics of this project, but my impression was that the predicted numbers were a bit optimistic.

Very good points about learning (two points, actually.) First, putting things off until the future makes the worth less, although you get to do them better. Second, we may have to actually do things in order to learn about whether those kinds of things work.

These are the same points that randomized experiment people make in development economics. But that's a situation where it is unproblematic to learn from the results of a policy. The whole policy is designed to be learned from. In this case, the policy's effects are bound up with a million other factors affecting the economy of California and there's no counterfactual. So it's unclear if we'll learn anything at all about whether high-speed rail is useful from doing this.

But, I agree with your first point. It doesn't look like we're going to get better information about the value of high-speed rail any time soon. And it's not as if the technology is advancing at a rapid pace, so we should either do it now or not at all.