“Level of Service” is a staple of traffic engineering. Its definition, according to the Highway Capacity Manual, is as such: “Level of service (LOS) is a quality measure describing operational conditions within a traffic stream, generally in terms of such service measures as speed and travel time, freedom to maneuver, traffic interruptions, and comfort and convenience.” It sounds clinical, inoffensive, and with its A through F letter grades, it fixes itself easily in the public mind. Says the parent: What, little Johnny’s getting an F? Well, we need to make improvements! Says the engineer: This facility has a Level of Service F. Well, we need to make improvements!

The “level of service” designation is a descendant of the strand of efficiency-minded engineering that came out of the Progressive age, a time which, as Peter Norton notes in his book Fighting Traffic: The Dawn of the Motor Age in the American City an important transformation took place: “Once a public space for mixed uses, and ruled by informal customs, the street was then becoming a motor thoroughfare for the nearly exclusive use of fast vehicles—especially automobiles.” Engineers went from counting people on streets to vehicles; and the key, seminal traffic engineer Miller McClintock went from arguing that “street capacity can be increased effectively by regulating traffic” to, once he had accepted funding from Studebaker, arguing for the “greater provision of street area [for automobiles].”

LOS is a classic case by which a bland bit of technical jargon conceals an entire ideology: Namely, that the purpose of a street is to move as many vehicles as possible, as quickly as possible. There are any number of problems that have been identified with that thinking. For one, how are we to judge the highway in the photo above? An economist might view it favorably as a piece of public infrastructure being used almost to its full capacity. But to a traffic engineer, that highway earns an “F”: “Operation with very high delays and congestion. Volumes vary widely depending on downstream queue conditions.” But another problem is that LOS designations are often issued during the peak morning and afternoon periods, meaning that a highway that looks like a failing “F” facility may be underutilized much of the rest of the day. Of course, this “building a church for Easter Sunday” is a staple of highway and parking demand modeling; we hear plenty about the economic losses due to congestion but never seem to hear about the economic losses of overbuilt, under-funded highways that are well under capacity most of the day. Another problem with LOS, particularly for urban environments, is that its view of the supposed “service” is entirely monocultural. The aforementioned “comfort and convenience” refers to drivers’ ability to whisk unimpeded from point A to point B, but says nothing about how pedestrians or cyclists or people living along that street may find it.

As an example, look at the photo below, which comes from an excellent paper by Ronald Milam, of Fehr & peers.

The existing intersection shown has a LOS of “E.” (even though, you’ll notice, it looks hardly filled here). The new sections being added represent “improvements” that will bring it up to Level E. But now imagine you’re a pedestrian. Suddenly the width of the intersection has doubled, and a more or less iron law of traffic engineering is that pedestrian safety declines as streets get wider. A study I cite in Traffic has also shown that adding lanes at an intersection is a process of diminishing returns, with increasingly less bang for one’s buck — but this too is something that gets swept under the rug in that unimpeachable move from “E” (bad) to “C” (good).

The LOS, with its seemingly quantifiable (if hollow) authority, washes over any other number of considerations. As Milam writes, “Widening a roadway to maintain ‘acceptable’ traffic flow may involve removing homes, trees, or open space in some cases; things on which a community may place a higher value than travel time. However, formal mechanisms don’t generally exist in local policies or procedures to weigh these factors against each other, so the LOS threshold usually takes precedence.”

What set me off on this LOS tangent was a post at Fehr and Peers’ blog about how the state of California is considering doing away with LOS. The Governor’s Office for Planning and Research, which is currently updating California’s Environmental Quality Act, has long relied on LOS to measure transportation impacts. But things could be changing, the post noted:

Which brings us to January 2009. OPR has released proposed changes to the CEQA checklist which eliminate the above language, replacing it with language relating to the number of automobile trips or vehicle miles traveled (VMT) a particular “project” would generate…

…This proposal comes on the heels of the City of San Francisco’s proposal in the Fall of 2008 to do just the same thing. It also follows an LOS Forum on this topic that was held at OPR in December. If adopted, it would mean that automobile LOS, which describes the level of congestion and delay on a road, would be abandoned in favor of impacts being based on the amount of automobile travel generated, irrespective of roadway capacity.

What do you think? Is LOS itself getting an “F”? Are other states moving to overhaul this obsolete tool?

This entry was posted on Tuesday, March 3rd, 2009 at 8:50 am and is filed under Congestion, Traffic Engineering, Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.