OPINION: Transit has a negligible effect on traffic congestion

In 2019, I attended a community meeting where a Houston Metro representative spoke to campaign for its 2019 referendum to approve an additional $3.5 billion of debt to expand Metro’s footprint. The representative began her presentation by asking, “Who here enjoys sitting in traffic?” Most in the crowd moaned or shook their heads. She continued, “That’s right. None of us do. That is why we need to make sure the referendum is passed.” Most in the crowd nodded in agreement.

What was implicit in the spokesperson’s question, and was accepted by most in the audience, was the assumption that transit reduces traffic congestion. It is an assumption that is also widely accepted by the public. In my survey on transit issues, the reason that respondents most frequently mentioned as the most important rationale to have a transit system was to reduce traffic congestion. And it seems intuitive that if we can bunch up a lot of people and put them on a bus or train, we would take those people off the roads and thereby reduce congestion.

But is that intuition and the resulting assumption supported by the data?

In Houston, the answer is pretty simple. Transit has a negligible effect on traffic congestion here. That is readily apparent by looking at the percentage of daily trips that are made on transit. As I recently shared with you, Metro carries less than 1% of the daily trips made in the region. If Metro were to go away or by some miracle, say, double its ridership, few in our region would notice any difference in traffic congestion.

The conclusion is further supported when we look at transit’s specific effects on congestion.

To start, at-grade1 transit is not “congestion free.” Buses and trains using the same roadways at other vehicles contribute to overall congestion. And they do so disproportionately to other vehicles because of the size, difficulty in negotiating tight turns, and frequent stops. Think about the number of times time you have had to change lanes to go around a stopped bus. Also, buses in particular move very slowly. This FTA study (p. 17) found the average speed of a city bus was 8.8 mph.

We also overestimate the number of riders per transit vehicle. The typical city bus will carry about 50-60 people. But according to the same FTA study (p.21), the average number of riders on a bus at any given time about nine. I suspect that average is even lower in Houston. All of us have passed virtually empty buses and trains. So, transit does not “bunch up” as many people and get as many cars off the street as many assume.

Also transit trips are not very long. The FTA study (p. 18) also found that the average bus trip is 3.7 miles and the average light rail trip is 5.2 miles. Since Houston is more spread out than most cities, our average is likely somewhat higher here. But the reality is that most transit trips are relatively short, which further diminishes transit’s effect on congestion, particularly in a large spread-out city.

Finally, most of the traffic congestion that Houstonians face on a daily basis is on a freeway. However, only a small portion of Metro’s service has any effect on freeway congestion. Its service is primarily concentrated on city streets not the freeways. The only Metro service that directly effects freeway traffic is its Park & Ride service. But the total daily ridership on the Park & Ride system is only about 16,000. According to TxDOT’s traffic counts, Houston’s Freeways move something over 2 million vehicles each day.2 So, moving only something like .1% (one-ten of one percent) of the current freeway traffic has virtually no effect.

Certainly, there are some cities in the world that would be overwhelmed with traffic congestion without their transit systems. But those are very dense urban settings like Manhattan or Paris. However, all of those systems are grade separated. I am not aware of any at-grade system has ever been shown to have any significant impact on congestion.

And even in such cities, transit has not “solved” traffic congestion. If you don’t believe that, I suggest you try to drive into Paris on one of its freeways at rush hour.

But there are a few specific routes, at particular times of the day, that Metro probably does provide some congestion relief. One such area is around the Texas Medical Center.

Metro’s year-end ridership report suggests that about 6,000 people who work at the TMC park in remote lots south of TMC, along the Metro’s Red LRT and take the LRT to their jobs. Adding 6,000 more cars entering and leaving the TMC, especially during peak hours, would clearly make the already heavily congested area even worse. However, offsetting that relief is that the LRT system has made a mess of traffic along Fannin. The LRT probably should have been elevated through the TMC.

If you look through the ridership report on Metro’s bus routes (Metro Ridership Report at pp.3-4) there are about a dozen routes that might provide some relief from traffic congestion, such as Westheimer (Route 82) which has about 12,000 riders per day. But there are less than 20 boarding per hour for most of Metro’s bus routes. If you think about taking less than 20 cars per hour off most of these throughfares, the illusion that Metro is making any significant contribution to reducing traffic congestion evaporates pretty quickly.

Outside of areas where there is a very dense residence population or a very dense concentration of places of employment, transit will never have any significant effect on congestion. For better or worse, Houston has almost no such concentrations. And if we really want to do something about traffic congestion, there are many other strategies that are both more effective and don’t even come close to the obscene cost of transit. More on that subject soon.