Should the 70 Be Merged With the 9? Or the 73? Or Something Else?

One of the biggest changes recently proposed by TriMet is the idea of merging the 70 and the NE portion of the 9, creating a new crosstown route connecting Milwaukie, Sellwood/Westmoreland, the inner SE, Lloyd Center, and Alberta. There is map available here. The agency must have anticipated some pushback from riders of the 9 accustomed to a direct link to downtown, as they included this line at the end: “If lines 9 and 70 do not have their routes combined, an alternative may be to combine the Line 73 with Line 70.”

Combining the 70 and 73 intuitively makes sense, as they are both currently crosstown routes and doing so would not take away anyone else’s direct downtown service. As I discussed in my last post, the 73 is a poorly-designed route that attracts very little ridership compared to other routes in the NE. The 70 and 73 both make the largely pointless diversion west from Lloyd Center to Rose Quarter, a segment that could easily be deleted from both with little mobility loss.

So if that is the case, why not leave the 9 alone and combine the 70 and 73? The major problem with doing so is that TriMet is also proposing to cut evening and weekend service from the 73. This means that if the 70 and 73 were combined, the existing 70 would have to be cut as well, weakening a fairly successful line in the process. I’m not sure we want this new crosstown service if it barely runs often enough to be successful.

Let’s try to attach some numbers to this.

Route 70 runs 104 one-way trips per weekday, 76 trips on Saturday, and 42 trips on Sundays.

Route 9 runs 106 trips per weekday day, and Saturday and Sunday each have 52 trips.

Route 73 runs 68 trips per weekday, 28 trips on Saturday, and 26 trips on Sundays.

One thing that immediately jumps out is that the 9 and 70 are much closer in terms of current service levels. The 70 has more trips on Saturday and fewer on Sunday, but total trips are fairly close. This may be why TriMet planners saw these as compatible routes to combine. The 73, by contrast, has much lower service levels than the 70. The 70 runs about 50% more trips on weekdays and more than twice as many on Saturdays.

If this change were not part of a budget-cutting process, the opportunity would be there to combine the 70 and 73 and give the whole new crosstown the same service levels of the current 70. This would probably be justified by the extra ridership attracted by a more useful route. However, in the current situation the danger is that they would just drop the 70 to the 73’s service levels, decimating a currently useful and popular route in the process. This is especially the case given that cutting night and weekend service on the 73 is on the table.

To see if this is true, we have to see roughly how many service hours TriMet is trying to save with the proposed changes. This calculation will not be completely accurate, since I don’t know the layover times for each route, but it should be fine for comparison’s sake.

If the 70 and 9 are combined, some savings will come from deleted segments. The 70 will be deleted from Lloyd Center to Rose Quarter, saving about 5 minutes per trip. The 9 will be deleted from downtown to Lloyd Center, saving about 10 minutes per trip. Total savings will be about 159 hours per week. TriMet also wants to cut night and weekend service from the 73. This will end up cutting 12 trips per weekday and all trips on Saturday and Sunday, for a total of 114 trips per week. Each trip takes about 30 minutes, so savings amount to 57 hours. Total savings from all three routes is about 216 hours, so that is the number to beat in any alternative.

So what if the 70 and 73 are combined and the 9 stays the same? Well, we still save on the deleted Rose Quarter segment of the 70, plus we can delete the same segment of the 73. This totals to 86 hours, still 125 hours short of the previous scenario. So how much would have to be cut from the new 70/73? It looks like we would need to cut about 136 trips from current levels to make it pencil out. To do so would basically require running the new route at current 73 service levels, meaning barebones weekend service and about 50% fewer trips each weekday. That or they could completely cut night and weekend service. Such a low level of service will not be sufficient on a crosstown route that will connect so many dense neighborhoods and destinations. The 75 and 72, comparable crosstowns, are two of the most successful routes TriMet runs, with high frequencies still often unable to keep up with demand.

So what is the lesson from all this? I would say the lesson is that there are good reasons for both TriMet and the public to prefer connecting the 70 with the 9 rather than the 73. Since service levels are similar, riders of both the 70 and 9 will not see a drop in frequency. True, riders of the 9 headed to downtown will have to transfer to MAX, or the streetcar, or the 8, or the 12, or the 19, or the 20, or the 15, or the 14, or the 4, and so on. Maybe you can see why I have a lack of sympathy for this plight. The incredible abundance of transfer opportunities to frequent services will make this about the easiest transfer in the whole system, most likely under 5 minutes most of the time except in the evenings or on weekends. If the 9 is protected from any change and the 70 and 73 are combined instead, the result will be either the creation of a very-infrequent new crosstown service, cuts to frequency on the 9, or a combination of both. We need to be honest about these tradeoffs.

Some people have also suggested merging the 8 with the 70 instead, since those would be more of a straight shot north. When TriMet first designed the grid decades ago, they actually wanted to do this very thing, but they back down under pressure from the Irvington neighborhood, which didn’t want to lose their direct downtown service (sound familiar?). On the face of it this idea might not work in this situation, since the 8 has higher service levels than the 70, but it could work if structured correctly.

On that note, I would like to close with a not-so-modest proposal of my own: Move the NE portion of the 9 over to 33rd (with a connection to Concordia University), replacing the 73 with the 9, which would continue to go downtown. Merge the 8 and 70 to form a nearly perfect north-south crosstown. The portion of the 8 to Marquam Hill can be attached to the end of this new route on 33rd. By my calculations, doing this would save about the same amount of money as TriMet’s proposal, it would keep frequencies intact, and the 70 could actually increase in frequency. The Alberta and Irvington neighborhoods would have direct downtown service on each side (MLK and 33rd), with crosstown service in the middle on 15th, with the plethora of transfer opportunities available. If we’re going to have direct-to-downtown service anywhere in this area, 33rd also makes sense because it is in the middle of a large gap between MAX stations.

With this change there would be there would be a largish gap between the transit routes on 15th and 33rd, but that is not atypical in the Portland grid. This would also create a nice alternating sequence of downtown and crosstown routes, creating a more balanced system that doesn’t overly favor downtown-bound riders over crosstown riders. The 9’s route up Alameda Ridge has always been an oddity, running on tiny residential streets and switchbacking up the ridge at low speeds. I’m pretty sure frequencies are limited by the fact that the buses can’t even pass each other on these streets. There is a reason most buses run on main arterial roads–they are wider and faster, have more businesses and other destinations in walking distance, and are easier to site stops and shelters. The NE tail of the 9 was an admirable effort to force a grid onto an area, but it really doesn’t make sense when a perfectly good arterial with lots of business activity, NE 33rd Ave, is right nearby. It’s time to try something bold and different that can retain maximum mobility and access in the face of these unfortunate service cuts.

28 Responses to Should the 70 Be Merged With the 9? Or the 73? Or Something Else?