Elon Musk took a lot of crap on Twitter recently when he questioned “induced demand,” a popular theory of urban planning. While the theory does have its skeptics, it has been shown to be true across a wide variety of situations. That having been said, it’s also widely misunderstood among environmentalists and urban advocates on Twitter.
Or, as Inigo Montoya says in The Princess Bride: “You keep using that word. I do not think it means what you think it means.”
The basic concept is solid. On public roads without tolls, we’ve seen that increasing travel space with new roads and more lanes doesn’t get rid of congestion. Why? Because the wider road allows more traffic to flow, which attracts more traffic, ultimately leading to the congestion not improving after roads are expanded. The perceived cost of driving on the road, including things like time spent in traffic, lowers when the road widens. This lower cost attracts more traffic until the cost goes back up with new congestion.
Many commentators I’ve come across regard this as an unassailable truth, and apply it very widely. The problem, though, is that they often apply the theory to situations it wasn’t developed for, or assume it proves things that it doesn’t prove.
To really understand this, we have to look at what Induced Demand Theory doesn’t cover.
For one, the correlation of 1.03 found by Duranton and Turner (whose paper in the October 2011 American Economic Review revealed the modern understanding of induced demand), doesn’t say that there’s always a perfect 1-to-1 growth of traffic compared to freeway capacity. The national average was 1.03, but places did fall above and below that. Robert Poole at the Reason Foundation points out that cities which have invested more in road construction have a lower correlation — that is, the new additions did not result in the same saturation of traffic. While the problem of congestion isn’t totally solved, it does show that there are limits to the applicability of induced demand theory (as displayed in the upper part of the chart).
Another area where the Twitter experts often err is their assumption that public transit is the answer to congestion. That’s an assertion that Duranton and Turner said was false. “… we find no evidence that public transit affects VKT…” the paper says (VKT is vehicle kilometers traveled). They point out over and over that adding transit does not remove traffic from the roads in any meaningful way.
Another thing not factored in is that a congested wider road moves more vehicles than a congested narrow road. While the issue of congestion is not solved, the extra trips induced that re-crowd the road are still happening and still have benefits to society. Additional economic activity is happening. While there’s plenty of room for debate over whether these benefits outweigh other costs (pollution, roadway fatalities, etc), it is something that does need to be factored in.
Finally, the data supporting induced demand is only for freeways. Toll roads don’t work the same way, because the economics are different. While the perceived cost of being stuck in traffic is a factor on freeways, toll roads add an additional charge that may vary with time of day, allowing pricing to reflect supply and demand. These road additions can relieve congestion without inducing too much new demand as to negate the benefit (the lower part of the chart in this article explains this visually).
This last point is where we really find that induced demand doesn’t apply to The Boring Company’s planned tunnels. Going in the extra roads privately supplied in the tunnels will not be free. Private toll roads simply haven’t been shown to contribute to induced demand, so to apply this to Musk’s idea is absurd.
According to available data, the worst thing that could happen with The Boring Company tunnels is that they could become a “Lexus lane” for rich people to avoid traffic. That only occurs if the tolls are too expensive. In all other circumstances other than giving away access for free, they will not contribute to induced demand.