Cox argues that denser metro areas also have more traffic. He argues that part of the reason is their unwillingness to build new highways and that less traffic can be an economic advantage for an area.
By Cox's reasoning, low-density Richmond should have a huge economic advantage since it has the least congestion of any major metro area. Hampton Roads (which is actually even less dense than Richmond) doesn't fare quite as well but is still on the low end for traffic.
There are a couple of small problems even with this graph. Bacon points out one - Cox fails to take into account that larger metro areas simply have more traffic regardless of their density - and that they also tend to get denser as more people try to stay connected to the core of the area. Another is that he looks only at the delay during peak periods, which tells us how bad rush hour can be but isn't always a good metric for traffic in general.
But these are both minor points. The real problem with Cox's analysis is that he makes a point that is correct and obvious, then draws or infers conclusions that are completely incorrect. More density does, of course, mean more congestion. If there are more people in a given area, there will be more of them going places and doing things. And just building new transit or making areas walkable does not mean that there will be less visible traffic on the roads. Because roads are free and convenient, people tend to max them out before shifting to other modes. But more visible traffic is not a sign that density is dysfunctional, nor is it always a bad thing for cities, for several reasons.
First, he measures traffic congestion rather than access. The two are very, very different. Most people perceive the inconvenience of traffic in terms of how fast they can drive on a road, which is ridiculous. They ought to evaluate it in terms of their increased or decreased access to possible destinations. So yes, ten minutes of driving in Manhattan might barely get you a mile. But that mile driving radius gets you access to a million people, several million jobs, and tens of thousands of retail stores and restaurants. Contrast that with suburban Richmond. Ten minutes of driving in Chesterfield County might get you geographically farther in any one direction (including time on side streets to get to destinations), but that only gives you access to less than a hundred thousand people, and not nearly the concentration of jobs or amenities. Cities create value by being markets and markets that can provide access to more people result in more value creation. Part of that access also involves being able to use different modes based on the different circumstances of the trip. This mapping application, which is fantastic, gives a better idea of how much access cities are creating for their people and how traffic affects that: http://www.flaviogortana.com/isoscope/.
Here's a ten minute drive in Manhattan, covering an area inhabited by nearly a million people, with millions more jobs:
vs. Chesterfield County, where a ten minute drive gets you access to less than a hundred thousand people and jobs:
The same argument crops up between fire departments and planning departments. Fire departments, worried about their response times, often push for large roads and unsafe pedestrian infrastructure that pushes people further out. Even though they can't travel as quickly in denser cities, the concentration of people means more fire stations. In central Boston, response times are a fourth of what they are in the outer suburbs. Even here in Charlottesville, city stations can respond in less than half the time that county stations can, their complaints about narrow streets and tight corners notwithstanding.
Second, Cox fails to mention that, while traffic may correlate somewhat with density, it correlates even better with economic growth or decline. And that sort of invalidates his argument about a lack of traffic being a major draw. It is a draw – in the same sense that an empty room at a party is attractive if one room is overcrowded. But the whole reason you are at the party is to talk to people, so there is a reason the most crowded room tends to get more crowded over time until a group can form that’s large enough to spin off another conversation. The areas with the most traffic also tend to be the areas with the highest median incomes. Traffic is a sign that people have things to do and places to go. When the economy tanks and unemployment rises, traffic tends to get lighter.
For years, planners and economists assumed a pretty tight relationship between vehicle miles traveled and GDP growth, and they were largely justified. In the last five years, however, the situation has started to change.
Third, from a fiscal (and to a certain extent an environmental) perspective, more traffic, if it can be controlled by traffic management systems, means more efficient use of the roads. This is generally true at the metro area level that Cox is analyzing since most roads are far below capacity most of the time (it’s not necessarily true for a particular road after it reaches the congestion point). Areas that have very low traffic congestion are probably paying far too much for their roads given their (probably poor) economic power.
I will use this as yet another excuse to post my comparison of the amount of pavement per vehicle in Charlottesville vs the pavement per vehicle in suburban Albemarle County:
