Response to Comment on Transit Pricing

An earlier post advocating peak period and distance-based pricing on transit prompted a thoughtful comment. Here's my response to the comment:

• Transit has often shied away from peak pricing for a number of valid reasons, mostly because the public does not view transit as a separate market from roads. If you think transit is a congestion relief mechanism for the road system, and that transit should be subsidized according to the congestion relief benefits it provides, then you face the peculiar situation where transit fares should actually go down as road congestion goes up. Even if transit ridership is flat, the value of that ridership can increase purely because road congestion worsens.

A: Transit is in part a congestion relief mechanism. But that said, peak pricing for transit can increase the efficiency of the transit system as a congestion relief tool. Peak pricing during the rush hour period encourages people to defer less important trips to the non-peak period. By reducing demand somewhat during the rush hour periods transit systems can run more smoothly during those peak periods, as anyone whose trip has been delayed by people fighting to get on and off packed vehicles will attest. Transit is a more viable alternative to the private car for work trips to the extent it can operate more smoothly and with fewer delays.

• If peak pricing on roadways ever really happens, transit will be seen as the fallback alternative to those who are priced off of the roads. Establishing peak pricing on transit in that situation will be a nigh impossible political sell. In fact, a peak transit discount might be the only way to package road congestion pricing.

A: Giving away highway space may have been a great social good (if you ignore environmental consequences) when there was ample capacity. That pricing strategy is counterproductive today, when demand for road space is outstripping supply. Electronic toll collection systems (e.g., I-PASS) have advanced sufficiently that highway tolling can be done without forcing people to stop at tollbooths. London and other cities that have implemented highway pricing systems point the way.

In this region highway pricing could be implemented relatively easily, at least from a technical perspective, using a few key points on the expressways as collection points. Alternatively, the bridges over the Chicago, Des Plaines and Fox Rivers could be tolled. The revenue generated could subsidize transit service and perhaps transit fares could be kept steady or actually reduced in peak periods.

There is a better approach, however, in a road pricing environment that would enhance the efficiency of the transit system and address the equity concerns that you mention. That is to have peak period pricing on both the highway and transit systems. This peak pricing would help push non-essential highway and transit trips to the non-peak periods. At the same time, the electronic toll collection and transit fare card accounts provide a way to target rewards to low-income transit users or all users who travel on transit in particularly congested corridors.

• Peak pricing will indeed encourage peak shifting from on- to off-peak, but the peak price will fall heaviest on those who have no flexibility in their arrival and departure times – highway or transit. To the extent that lower-income people have less flexible hours for jobs, this could be socially destructive and is the impetus behind sloganeering about "Lexus lanes." By contrast, some researchers think that lower-income people will actually benefit from congestion pricing by avoiding late fees from day care, etc.

A: It is hard to generalize who has more flexibility in hours, the rich or the poor. I believe there is credible research concerning high occupancy toll lanes in San Diego that finds a high level of satisfaction across income levels.

It is also important to recognize that one key benefit of congestion pricing is faster and more reliable transit service. This is a direct benefit to public transit users, who on the whole tend to have lower incomes than those who use automobiles.

• Transit “peak” pricing ideally should not be based on distance traveled (à la WMATA). Theoretically, transit should be priced at the point where the marginal operating or capital costs are actually incurred by the agency – the peak load point on a rail line or bus route. For example, say that the CTA’s Red Line’s peak load point in the AM is southbound at Grand station – the CTA sets their peak-of-the-peak schedules, and plans their fleets and signal systems based on the frequency required to meet that load. The passenger that boards at Howard and gets off at Addison at 8:15am has a marginal cost to the CTA of zero. In fact, only the passengers onboard the train at Grand should be allocated peak costs, because if the load at Grand were to decrease, the CTA could save a train. In serving the peak load point at Grand, the CTA would be providing the Howard-to-Addison capacity anyways, so that passenger should pay no peak charge. However, explaining true marginal cost pricing like this to normal passengers would be exceedingly difficult. Distance-based fares are a much easier sell. My point is only that peak pricing on transit isn’t as clear-cut as peak pricing on roads, and that transit’s substitutability to auto travel might actually require the opposite of peak pricing in a political package.
  

A: I'm not sure I understand the difference. In your example the heavy load of Grand station passengers challenges the capacity of the transit system as surely as a choke point in an expressway (e.g., Eisenhower from Oak Park to I-88). One potential advantage of congestion pricing on highways is the ability to set up a toll point at the choke point and use variable pricing to maintain traffic flow.

Given the relatively high cost of providing transit service, I'm not sure that there is much political will to "require the opposite of peak pricing in a political package" involving transit. The Auditor General's recent report (pg. 10) indicates that it costs the taxpayers about $2 per transit trip in the six-county RTA area in operating subsidies alone. On the capital side, transit eats up a much larger share of public capital dollars than the highways relative to its market share.

• Last, is it really fair to ask transit riders to pay more efficient prices and leave prices so misguided and distorted on the auto side? I say, fix the price of auto travel first – this would stimulate demand for transit-friendly urban development, and who knows, as you say, we might even get the revenues from a road congestion pricing scheme dedicated to building more transit.

A: I'm not so sure that prices are "misguided and distorted" on the auto side. Private auto owners already are bearing much of the cost of the highway system. They pay the cost of owning and operating their vehicles, including the fuel taxes that fund road construction and maintenance. Lets say that these costs are $.50 a mile. Thus, drivers who take a 4 mile trip are paying the same as the straphanger who pays $2.00 to ride the bus the same distance. That same transit rider also gets a $2.00 subsidy. Maybe transit pricing is just as "misguided and distorted" as auto pricing in that transit riders pay far less than the full cost of their rides.

Even though drivers already pay a significant price for auto travel demand for highway space exceeds supply because auto travel is perceived to be much superior to travel by public transit by the vast majority of people for the vast majority of trips. I agree that there is room for using congestion pricing to better manage that demand and perhaps generate revenue for other transportation projects. I also agree that higher highway prices plus decent alternatives to auto travel--transit, walking, biking--can lead to a virtuous cycle that will reduce our dependence on the auto at least a bit in at least some areas.

However, your assumptions that auto owners somehow pay less than transit riders or that transit pricing is "efficient" while auto pricing is "distorted" likely are incorrect. Transit pricing would be more efficient if it were distance based and used peak period pricing. Moreover, the strong demand for rail travel and the continuing erosion of the market for bus travel in the region suggest that rail and bus trips should be priced differently.