Friday, February 17, 2012

What is congestion?

A lot of people really care about congestion. A major source of capital funding in this country is the federal Congestion Management and Air Quality grants. Occasionally, in an argument over some aspect of transportation policy, someone will announce, "But your proposal would cause congestion!" and sit back triumphantly, thinking they've just won.

In the past I've argued for congestion pricing, but honestly I just can't get too excited about fighting congestion. If you read my list of goals above, you'll notice that reducing congestion is not on there. Congestion does affect my goals, so I do care on some level, but other factors may outweigh it.

Unfortunately, "congestion is bad" has become such an article of faith that in 2006 when Mayor Bloomberg said "We like traffic, it means economic activity, it means people coming here," he was attacked by both car activists and cycling activists. I hope that you, my readers, will be a bit more tolerant.

But what is congestion? I think it's important to distinguish a few different aspects. First of all, congestion is not confined to the movement of vehicles or even people. We can have literal congested arteries, where fat deposits slow or block the flow of blood. We can have sinus and chest congestion, where excess mucus can interfere with our oxygen supply or our elimination of carbon dioxide. There is definitely congestion on electronic data networks, just this evening I heard a podcast where someone was talking about congestion in the power grid, and I imagine there's also congestion in aqueducts and sewers.

Back in the realm of vehicles, the most familiar form is road congestion, but we also have air traffic congestion and rail congestion. Road congestion isn't always vehicles: we can have congestion of bikes, pedestrians, roller skaters or golf carts.

What all these things have in common is trouble with flow. Interestingly, it's not all about speed. Sometimes the overall throughput is satisfactory, but the trouble is that there are stops and starts. Other times the flow is constant but it's too slow. There are also two different sources of congestion: either it comes from increased demand, or from other stuff blocking the duct. For example, in power congestion it's simply that people are trying to force more electrons through a grid of a constant size. In contrast, nasal passages and bronchial tubes stay the same width, but get blocked by mucus. Sudafed works by decreasing the production of mucus, while Mucinex works by thinning the mucus; both reduce the blockage, allowing air to flow better.

We see the same divisions in road congestion. Sometimes people get to where they're going in a reasonable amount of time, but they spend too much time stopping and starting, while other times the cars flow at a steady pace but slowly. Sometimes the traffic pattern induces a large number of cars through a road that isn't built to handle them, while other times it's a whole bunch of low-priority vehicles that block the higher-priority vehicles.

For a concrete example, we can turn to Mike Grynbaum's report from last year about the large number of off-duty taxis heading from Manhattan to Queens and back at 4PM for the change of shift. Just as it's more important for air to flow through our noses and lungs into our bloodstream and back out than for mucus to flow out, it's more important for buses, cars and trucks to get to Queens and back than taxis.

It's not that you don't need to drain mucus from your sinuses, and it's not that the taxis don't need to get to get out to Queens for their change of shift. It's just that the mucus could drain slower and leave more room for you to breathe, and the taxis could travel slower and leave more room for other vehicles. The same is true for all the entitled drivers who could be taking the train or traveling at less congested times.

So there you have it: off-duty taxis and discretionary car trips are the snot of New York's roads. Parking pricing is the Sudafed for that snot, and congestion pricing is Mucinex.


arcady said...

Keep in mind that congestion is also often considered to be delay relative to driving on an empty road. So if the average commuter drives 20 minutes on a road where the uncongested travel time is 10 minutes, that's congestion, and that's bad. If the average commuter instead drives 40 minutes at 70 mph, there's no congestion and everything is perfectly fine.

Zoltán said...

Here's the problem from where I stand: Congestion (at least with PT in exclusive lanes or with grade separation) is great from a perspective of getting people out of their cars, as it makes driving really inconvenient. But slow, stop-start traffic is terrible in terms of local air pollution.

Pricing, of course, allows for free-flowing roads while discouraging use, but it's politically difficult. Also, giving lovely free-flowing roads to the rich bothers me a bit, though I suppose of necessity the transit alternative should in that scenario making driving merely frivolous.

Pat said...

It seems to me that "reducing pollution" and "increasing efficiency" are both relevant to congestion, even if reducing congestion isn't a goal unto itself. Any solution to congestion is going to have net benefits along those lines, as long as the solution isn't "build more roads."