Sunday, November 30, 2008

How the Rivercrossing Project fails Transit

Transit is sick in Salem. We all know the Cherriots levy failed. But over the last year the Rivercrossing project has also gone down a path that denies the value of transit.

A recent memo from this past summer makes clear what many of us have supposed:
the Salem River Crossing project Build alternatives are designed to improve single occupant vehicle travel time…
There is no serious attempt to improve the movement of people and goods across the river in a way that is efficient, cost-effective, and anticipates both climate change and peak oil. The project is wholly business-as-usual.

Bicycle advocates have argued that the Governor’s climate change initiative, House Bill 3543, requires looking at ways to reduce greenhouse gas emissions. Building a bridge for single-occupant vehicles will only increase greenhouse gas emissions. The plan to build this bridge is patently inconsistent with the need to reduce emissions.

Bridge boosters claim that they have investigated the alternatives, and that the alternatives are woefully inadequate, offering small incremental improvements at best.

When we look at the analysis, however, it appears that boosters have stacked the deck against other modes and created something of a straw man argument.

First let’s look at congestion pricing and tolling. From the August 15, 2007 TSM/TDM (Transit and Roadway Efficiency) Concept - Analysis and Results memo (tsm/tdm = transportation system management and transportation demand management):
For modeling purposes, parking fees in downtown Salem were used as a one possible example of user fees that could be adjusted to influence travel demand. (2)
Even if we suppose that the model is perfect, the analysis didn’t try tolling or congestion pricing. It adjusted parking pricing only, and limited that adjustment to the downtown area only, penalizing those who wished to patronize downtown merchants, whose patrons currently enjoy free parking, and didn’t impose any additional costs on those who might be shopping on Lancaster or taking trips elsewhere in the city. Although the modelers don’t discuss this explicitly in the memo, it’s likely that the parking fees didn’t reduce trip demand so much as shift discretionary trips from downtown to elsewhere. The minimal reduction is probably also an admission that many of the trips don't end or begin in downtown.

They conclude saying,
The results show that increasing the area where parking charges are applied in the downtown area and tripling the average amount charged could reduce auto trips by 10% (from 86.6% of trips to 76.2% of trips). (3)
Now let’s look at transit. Here the assumptions are even more narrow.
The transit element consists of the operation of two new routes, one on Highway 22 and one on Wallace Road, each serving downtown (and possibly other employment destinations in the city)…. Stops were also minimized or limited to the park and ride lots provided. (4)
The model added only two new stops, at the park and ride lots! The model did not add a robust level of enhanced transit service with a network of stops near the front doors of residents. The transit service is minimal and inconvenient, and makes people get in their cars – why would they get out of their cars to stop at a park and ride lot for a short trip into downtown? This is a fatal simplification.

Moreover, the analysis only looks at getting people from the park and ride lots to downtown. It supposes they don’t have to transfer to get elsewhere in the city. Eliminating transfers is second fatal simplification in the model.

If you want people to use transit, it should be direct, frequent, and easy to use. The level of transit service used in the model fails two of the three.

Nevertheless, the analysts believe otherwise:
The level of new transit service assumed in the model is a high level of service for the routes and geographies served. However, as shown above, this level of service does not produce a significant shift in total trips across the bridge. The bridge users in the peak hour have trip origins and destinations that extend well beyond the area served by the two new transit lines. (6)
Note that they understand the problem of transfers, and yet insist on ignoring it.

Not surprisingly the analysis concludes:
This level of additional transit service would result in only very small daily changes in transit ridership compared to increasing the parking charges....the model shows that adding the increased transit service does not result in additional reductions in vehicle trips across the existing bridges.(5-6)
In order to knock down the extent to which virtually any TSM/TDM option, like congestion pricing, tolling, or transit might relieve congestion in a no-build or low-build option, analysts used two poor straw men, that of parking pricing and inconvenient transit, to argue that the whole suite of TSM/TDM strategies cannot succeed.

This is confirmed in the July 21, 2008 memo on the TSM/TDM/Transit Expanded Subcommittee:
Analysis to date has shown that providing transit system and operations improvements in the absence of user fees (tolls or parking charges) to discourage auto trips results in small reductions in demand. (1-2)
As was demonstrated in the study of a stand-alone TSM/TDM alternative (see “TSM/TDM (Transit and Roadway Efficiency) Concept – Analysis and Results” memo, dated 8/15/07), improvements to the transit system such as more buses and improved headways still leave the transit system at a significant disadvantage compared to single occupant vehicles with respect to trip time across the river and individuals’ understood time value of money. Consequently, such improvements to the transit system will have relatively little effect on peak hour demand on the existing bridges until or unless there is a significant increase in travel time (e.g., significant congestion for single occupant vehicles but not for transit) or the cost of driving (e.g., tolls, congestion pricing, parking pricing, fuel prices) for single occupant vehicles. Because the Salem River Crossing project Build alternatives are designed to improve single occupant vehicle travel time over the future No Build condition and do not propose differential user fees (e.g., congestion pricing), a significant time vs. money disadvantage for transit trips would be maintained, even with significantly improved bus headways. (2-3)
Unfortunately, it appears the analysis was not designed really to challenge or even seriously to investigate the assumption that TSM/TDM options are powerless to reduce congestion and improve the movement of people, goods, and services across the river. It was inadequate and served only to confirm the prevailing assumptions about the values of single-occupant vehicles and their easy movement.

Other project documents are available on the Rivercrossing library page.

Important ones include:
The 1999 Oregon Highway Plan Policy Element, which includes the "mobility standard"
Traffic Modeling information from SKAT/MWVCOG
River Crossing Purpose and Need statement
Threshold Memo
Evaluation Framework Memo
Alternative Methodology Report
Alternative Evaluation Results Memo


MWVBTA said...

Here's an August 15, 2007 memo on a Task Force Meeting, which includes Glen Hadley's comments (and those of others) on the inadequacy of the TDM/TSM/Transit modeling & analysis.

MWVBTA said...

The Willamette River Crossing Capacity Study Executive Summary (June 2002) has this to say about transit:
"In 2015, the p.m. peak-hour traffic demand on the Marion Street Bridge is projected to be about 6,700 vehicles. The peak-hour capacity of the bridge will still be 4,500 vehicles resulting in a capacity deficiency of 2,200 vehicles. It is highly unlikely that any single strategy, by itself, will be successful in removing the excess 2,200 vehicles from the peak-hour demand. For instance, it would require about 44 buses (each with 50 sitting and standing passengers) during the peak hour to carry the 2,200 persons. To expect such a dramatic shift in transportation mode choice is neither reasonable nor useful."