Are there too many roads?

With today’s news that President Obama is coupling tax reform with a renewed call for a major infrastructure investment (following up on his fix-it-first proposal in the State of the Union Address), we should ask the Goldilocks question: is our road infrastructure too little, just right, or not enough? Of course this depends on how, there is not simply a “lump of roads”, but rather roads of particular designs, connecting particular origins and destinations, with a length, width, and depth.

This discussion is US-centric, but applies to many developed countries. Developing countries are not there yet.

Too Little:

  1. On the side of “too little” are some economists, following on the heels of David Aschauer’s work in the 1990s about the infrastructure investment shortfall. The claim is that infrastructure investments drive economic growth.
  2. There are arguments about economies of agglomeration, which infrastructure may enhance, by operating through accessibility. To the extent accessibility increases, agglomeration increases too. This is likely to be non-linear, but may be increasing or decreasing in a particular range.
  3. There is also the stimulus argument, from macro-economists looking for useful jobs projects, to employ the unemployed.
  4. Having some redundancy in the network is also valuable, as it makes the network more robust to failures (such as terrorist acts, bridge collapses, or superstorms).
  5. Stronger roads and bridges would enable buses and trucks to be more efficient by carrying more weight.

Too Much:

  1. On the side of “too much” are arguments like those in the “Strong Towns” movement who argue that we can’t afford to maintain the infrastructure we have.
  2. Also worth noting is the lack of economic development impacts of most new investments.
  3. History suggests that if demand has peaked, supply is probably also at the peak of what we can support. This was true of rail in the 1920s, at which point the mileage of rail networks (both intercity and urban) declined. This may now be true of roads, with peak travel occurring, as we see rural areas continuing to depopulate, some places considering gravelization, some urban freeways being taken down and not replaced.
  4. It is clear we under-price what we have, so naturally that leads to more consumption than if we properly priced things. It is also clear the roads sector (and some other transport sectors) are in the mature phase of development, and roads have at best diminishing returns on investment, if not zero or negative returns.
  5. It is also clear that most roads are mostly empty most of the time, and that we build many roads far wider than are needed, so wide they can be used to store cars 24 hours a day.

I would conclude, that in general, we do have too much road infrastructure.

We mostly have too much width, perhaps a bit too much length (e.g. we could reduce the rural grid from a 1 mile to a 2 mile spacing in many places), and generally not enough depth (pavement thickness).

Some of the excess width and length should be gracefully abandoned. It may not be worth demolishing, but entropy can have its course. When it is to be rebuilt, it can also be appropriately rescaled. The term road diet is sometimes used, and applies in some cases.

That does not mean that we have all the infrastructure we need. There are surely some investments that have benefits that exceed their costs.

It also doesn’t mean that all infrastructure should be abandoned. Most infrastructure is quite useful, and we have developed land conditioned on the existence of that infrastructure. So abandonment is not necessarily practical.

So in contrast with the Betteridge’s Law of Headlines that any question in an article title is answered with a “no”. Here I think the answer is generally yes.