Transit Utilization and Traffic Congestion: Is There a Connection?

[cpzilliacus]

Reason's Out of Control: Transit Utilization and Traffic Congestion: Is There a Connection? - Research finds no statistically significant evidence linking increases in transit utilization to reductions in congestion

• Statistical analysis of the 74 largest urbanized areas in the U.S. over a 26-year period suggests that increasing transit utilization does not lead to a reduction in traffic congestion; nor does decreasing transit utilization lead to an increase in traffic congestion.
• Policies designed to promote transit utilization can in certain instances increase traffic congestion–as appears to have been the case in Portland, Oregon.
• Vehicle-miles traveled per freeway lane-mile is strongly correlated with traffic congestion: the more people drive relative to available freeway capacity, the worse congestion gets.
• Data from New York and Los Angeles indicate that the most effective way to increase transit utilization is by reducing fares, as well as by improving basic, pre-existing service.

[Brandon]

Gee, that sounds like common sense.

[NE2]

There is for those who take advantage of the transit. Those who don't are part of the problem.

[Duke87]

Statistical analysis of the 74 largest urbanized areas in the U.S. over a 26-year period suggests that increasing transit utilization does not lead to a reduction in traffic congestion; nor does decreasing transit utilization lead to an increase in traffic congestion.

Two words: latent demand.

Additional transportation capacity of any sort (be it transit or roads) will not appreciably reduce congestion until all the latent demand is met. Only once you have capacity that actually exceeds all potential demand will the congestion go away.

[jeffandnicole]

There is for those who take advantage of the transit. Those who don't are part of the problem.

Two words: latent demand.

Responding to both results in this: In most cases, adding mass transit also adds new homes/condos/apartments/businesses along the route.  Unless every single one of those people will take the train, they are simply adding more people to the existing road network, without expanding the road network. 

In many cases, people taking the new mass transit system are those that already took mass transit.  Instead of people taking the bus, they are now taking the train.  The net effect on reducing congestion is 0, because the bus routes still exist.

And within the train ridership are those taking the reverse route, which must of the time doesn't involve congestion, so there's no congestion to reduce.

[cpzilliacus]

Two words: latent demand.

Additional transportation capacity of any sort (be it transit or roads) will not appreciably reduce congestion until all the latent demand is met. Only once you have capacity that actually exceeds all potential demand will the congestion go away.

Agreed.  Though I also think that we could manage that demand better with pricing of scarce peak period highway capacity.  And transit capacity.

[english si]

Agreed.  Though I also think that we could manage that demand better with pricing of scarce peak period highway capacity.

Demand pricing would simply aggrieve motorists who are already fed up of jams without having to pay for the 'privilege'. No one wants to drive on congested highways if they can avoid it, so trying to price them off makes no sense.

Demand pricing will not manage demand, other than perhaps, if just main roads were priced, push them onto inadequate rat runs, with all the assorted reasons why that is a bad idea.

And transit capacity.

Do you not do peak fares? That said, in London, it is more the case that off-peak fares exist to try and get people travelling in those hours, rather than peak fares (of over twice the cost) being some sort of successful demand management.

And the reason why our rail fares in the UK are regulated is so that they can't charge peak passengers overly massive amounts knowing that the passengers don't really have a choice (peak fares are really high in the UK anyway, not that it has successfully smoothed out demand, or stopped the trains being full in the peak hours)

[Duke87]

Do you not do peak fares?

Unless something has changed since I last rode it (which, admittedly, was in 2007), the DC Metro does have slightly reduced off-peak fares. It's not enough to be a noticeable difference, all it does is leave you as a tourist asking "why does my farecard have 10 cents left on it when I purchased the exact amount the chart said I needed?". And don't get me started on the fare scheme itself.

The subways and buses in New York do not do peak fares but Metro-North and LIRR do (NJ Transit does not).


Anyways, special peak hour pricing on transit is counteractive to the goal here: if you want to use money to deal with peak hour congestion, you should make rush hour highway tolls higher but make rush hour transit fares lower - the point is to get commuters out of their cars.

Of course, this scheme only really works if you've got highway congestion at the same time that you have spare transit capacity. In cities where both roads and transit are at or over capacity during rush hour, there's no pricing your way out of congestion, you can only build your way out of it.

[PHLBOS]

In Greater Philly, SEPTA's Regional (Commuter) Rail has peak & off-peak prices for its tickets.

[cpzilliacus]

Agreed.  Though I also think that we could manage that demand better with pricing of scarce peak period highway capacity.

Demand pricing would simply aggrieve motorists who are already fed up of jams without having to pay for the 'privilege'. No one wants to drive on congested highways if they can avoid it, so trying to price them off makes no sense.

I disagree, for this reason.  If a highway is severely congested, managing demand through pricing will increase the number of vehicles that can use it.  Seems counter-intuitive to some laypeople, but not to persons that understand highway capacity.  The best case I am aware of is Ca. 91 (Riverside Freeway) in Orange County.  There are four "free" lanes in each direction, and two managed (priced) lanes each way.  published reports have shown that the two managed lanes carry more traffic than the four adjacent "free" lanes.

Demand pricing will not manage demand, other than perhaps, if just main roads were priced, push them onto inadequate rat runs, with all the assorted reasons why that is a bad idea.

And transit capacity.

Do you not do peak fares? That said, in London, it is more the case that off-peak fares exist to try and get people travelling in those hours, rather than peak fares (of over twice the cost) being some sort of successful demand management.

There's not much in the way of alternative routing for someone that wants to use the Ca. 91 corridor (begin here eastbound, toll point here and end here).

As for pricing transit higher during periods of high demand, I agree with that.  I even feel that a "peak of the peak" for the peak hour of transit demand should be charged.

The Washington Metro got rid of "peak of the peak" fares when rides vociferously objected to same.

And the reason why our rail fares in the UK are regulated is so that they can't charge peak passengers overly massive amounts knowing that the passengers don't really have a choice (peak fares are really high in the UK anyway, not that it has successfully smoothed out demand, or stopped the trains being full in the peak hours)

I cannot speak to UK train fares, as I have never been on a long haul train there (only short trains in and near London).