Earlier this month, the Amalgamated Transit Union (ATU) has published a paper titled “The False Promise of Microtransit,” (pdf) a critique of flexible transit services like King County Metro’s “Via to Transit” and Pierce Transit’s “Runner“.
From the ATU’s press release:
“The reality of microtransit, like other technology fads, is that it’s less impressive and more expensive than proponents suggest,” said ATU International President John Costa. “It doesn’t reliably deliver cost savings, real environmental benefits, or equitable service. What microtransit does do is threaten good transit jobs – real, long-term careers – by shifting service to gig workers and setting off a race to the bottom. Nobody wants to talk about that.”
The paper describes four shortcomings of microtransit service:
Microtransit cannot efficiently scale to meet increased customer demand.
It has been shown to serve a younger, more affluent, and less diverse ridership than fixed route service.
Its environmental benefit is doubtful.
It encourages cost cutting through privatization and the degradation of transit jobs.
After discussing each of these shortcomings in detail, the ATU makes the following recommendations to improve transit without implementing microtransit services:
Investments should be made in the transit workforce to overcome the operator shortage and guarantee reliable service. Creating and sustaining good transit jobs is the best way to attract more workers to the industry. Good transit jobs benefit local communities, unlike microtransit companies that profit by misclassifying their workforce as independent contractors.
Bus routes should be reconfigured to better serve passengers when and where they need to travel. The reality is that commuting patterns have fundamentally changed. Transit agencies must also make changes to improve existing services, such as reducing headways, building better bus shelters, and enhancing pedestrian connections to stops.
The paper recognizes that microtransit may be used to meet coverage goals that would otherwise be unaffordable, such as service to low-population exurban and rural communities. In that case, the ATU makes the obvious recommendation that microtransit services be provided by public transit employees instead of independent contractors.
Transit planning consultant Jarret Walker has written extensively on the topic of microtransit, coming to the conclusion that although microtransit (also referred to as “flexible transit”, “demand responsive transit”, or “on-demand transit”), is much less efficient than typical fixed-route transit. Unlike typical mass transit, which becomes more efficient with increased ridership, microtransit becomes less efficient with increased interest. However, Walker concedes that flexible transit may be the most efficient way to provide coverage (but not capacity) to very low density or otherwise low-ridership areas.
The transition of underperforming fixed-route service to microtransit has been ongoing in the Puget Sound, most recently with Pierce Transit’s “retirement” of routes 13, 63, and 425, each replaced by Runner microtransit service. Only time will tell if these changes are equitable or efficient.
It feels that micro transit is often subsidized, poor quality taxi service geared towards local residents. Trying to take micro transit as the final leg to reach a destination is rarely an easy process. This makes it something like a park and ride in being mostly one-directional.
Why not just subsidize a taxi/Uber/Lyft for all? This could also be an option for park and rides. For stations with park and rides, provide taxi service to locations within 5 miles of the park and ride lot for the same price of parking there. This would give those without cars or going to the area as a destination the same access as those driving.
Pierce Transit had a pilot program with Lyft about 10 years ago. IIRC, it was limited to picking up a rider in a specific zone (there was one in the Chambers Bay area that had previously been served by an eliminated portion of Route 53) and taking them to the nearest bus stop.
It also requires a smartphone and app, or at least some versions do. You can’t go to a bus stop at a known time and find a bus; you have to summon it.
Metro created several Metro Flex areas (previously called Via or other names), which are intended to provide coverage between fixed routes or that have no fixed route. Examples are Rainier Valley (between fixed routes), SeaTac (no route on 24th, or from the botanical garden/community center to TIB station), Issaquah, and Kent.
Pierce Transit converted several fixed routes to Runner service. An example is Ruston Way between downtown Tacoma, the waterfront parks, and the Ruston suburb/industrial area. Locals are reporting that vans get overcrowded and turn people away. PT needs more vans to fit the crowds, but that’s a significant expense and may not happen. It also requires a mobile app unless you look very carefully at an obscure part of the website and find the phone number to ask for a pickup. If you don’t have a phone, you can’t even call unless you can find a payphone (ha ha) or borrow somebody’s phone. So you can be right near a former bus stop and have no way to summon the bus.
Doesn’t seem very equitable now, does it?
Government run micro transit can’t compete economically with private micro transit because of labor and other costs, and the software will never match Uber’s. Most of the country has the Uber app. already on their phone with a credit card on file. Service is very fast. The number of miles Uber delivers each year continues to increase dramatically, and it is fair to assume those miles are taken from somewhere. It is a good complementary service to public transit and driving, although more expensive depending on number of riders, especially to longer routes like Link or rail or air. It also helps reduce the amount of parking including street parking that is necessary in urban areas where Uber does best and is most competitive.
Some can’t afford Uber, and some have disabilities that make micro transit necessary to get to transit, so there should be an on demand public service for them, although it often requires a reservation days in advance, and can be unreliable. I would be interested to know the costs of public micro transit to public transit vs. just funding an Uber from doorstep to final destination, especially for someone who is ill or disabled and who struggle with transfers or last mile access from transit.
I don’t think public transit dollars should go to Uber/Lyft in general, and clearly Uber doesn’t need public funding, although it may be cheaper for public transit agencies to contract with Uber for discounted services for micro transit for the disabled than implement its own services. That said the amount spent on park and rides is astronomical, but folks have to get to and from transit somehow. Not everyone can walk to transit, especially outside urban areas, and safety is sometimes an issue. Look at the cost per rider mile for buses in some of these outer areas like S. King Co. which can be more than $10/rider.
Personally, I think Metro’s expansion of micro transit, even in wealthy areas like Sammamish, is a sop to cities like Sammamish that get very little transit service but pay a lot of taxes but demand land use that makes transit service impossible. I also think Metro is concerned the future is micro transit based on the growth of Uber and Lyft, but those services really poach drivers of cars more than public transit, especially outside urban areas where Uber does best. Folks who can afford Uber generally don’t ride public transit.
The elephant in the room is driverless technology in the long term and how that reduces costs for Uber and transit, and in the short-term better implementation of a shared ride service that can combine riders living near one another going to the same destination, especially trips to the airport or from suburbia to urban areas like downtown Seattle so the cost is much lower.
Link benefits from Uber, so does long range rail and air travel, driving a car is reduced by Uber, and short urban trips on public transit suffer but a city needs less public/private parking. All can co-exist but Uber doesn’t need public funding for its role in transportation.
Public microtransit is usually just contracted with private (non-union) agencies. In a sense they do the the same thing as Uber/Lyft but there are differences. Uber and Lyft charge varying rates depending on the trip you take and when you take it. If it is a high demand time then they can charge a huge amount. If the cost gets too high then people just give up on it. But they also count on some part time drivers noticing the surge and decided to drive during that period.
Thus you really have no idea how much money will be spent on service. In contrast, with microtransit you have a limited number of buses, and they respond to passengers when they get to them. This keeps the cost the same every month, but it means that riders may have to wait a really long time for the bus. Basically the cost is kept relatively low because they put limits on the service. If they paid Uber/Lyft to give people rides the cost could be extremely high. There is nothing stopping a driver from charging a hundred bucks for a ride, and since the rider wouldn’t be paying it, the cost to taxpayers would be very high. I’m not saying they would purposely gouge the public agency, but it would just work that way. It would be like driving through variable tolls on a stolen credit card. The tolls aren’t set artificially high, you just don’t care.
Of course not all taxicab companies work like Uber and Lyft. Many have set rates that don’t vary based on the time of day. I could see an agency contracting with someone like Yellow Cab. Prices could still vary, but the costs would always be related to set pricing. I’m not sure it would be cheaper than what the companies provide now though.
Uber costs $10 for even a short trip. That adds up quickly if you make several trips a week or twice a day. Even middle-class people would find that hard to afford or justify. It’s not a substitute for regular transit that costs $2.75 and accepts inter-route transfers. Making it like that requires a large subsidy. That’s what Metro Flex and Runner do; they provide a large subsidy to keep the fare the same as buses. The subsidy is large no matter whether it’s in-house or outsourced to Uber/Lyft.
Some cities actually have replaced all their bus routes with subsidized Uber, so it’s not just theoretical. In one city that did this, lower-income residents said they can’t afford the higher fare, so they do without, and walk long distances to work or around inconvenient highways.
Uber is an investor-driven corporation, so they jack up the costs to maximize profits. Public transit operators just have to reach the union pay level (if they have unions), but there’s no cost-increase pressure beyond that.
” Most of the country has the Uber app. already on their phone with a credit card on file”
I doubt it’s above 50% or 75%. Even if it is, that leaves out people without credit cards or smartphones or unwilling to install the Uber app — a significant part of the public transit market.
But wait, it gets worse. The reason Jarrett Walker is so against microtransit is that it typically averages 2-3 passengers per hour, while even a minor fixed coverage route that “looks pretty empty” gets 10+ per hour. The vehicle sizes overlap, so the difference is the nature of fixed routes. A demand-response taxi has to travel empty to reach the next passenger, and loses time detouring to side streets for one person. That’s what makes the passengers/hour so low. A fix route has stops in a straight line where the highest ridership potential is.
You can see that with low-level coverage routes like the 906 (Southcenter, south/southeast Renton). I took it from Southcenter to Ikea on a Saturday afternoon, and counted 8 people before I got off. That’s only a third of the route, and the entire route from end to end is 45 minutes. So it could easily get 10 or even 20 passengers/hour. Another is the Snoqualmie Valley Shuttle. When I rode it one Spring weekday midday, it got 20 passengers between Snoqualmie and Duvall, and the end-to end time is 1 hour 15 minutes.
Both of those routes are semi-fixed, which gives the best of both worlds. Most of the stops are fixed, and it has a published schedule, but the schedule is padded to allow flexible areas at one or both ends. The Valley Shuttle seems to have 10 minutes extra each for North Bend and Duvall.
“I think Metro’s expansion of micro transit, even in wealthy areas like Sammamish, is a sop to cities like Sammamish that get very little transit service but pay a lot of taxes but demand land use that makes transit service impossible.”
The push is coming from car-minded people who don’t want to ride with strangers and want a vehicle to their driveway. Sammamish is full of people like that. Collectively they may pay disproportionate taxes because they’re rich, but they’re not that many people (because the area is low density), so they’re tiny percent of the county’s population. King County shouldn’t give them as much transit service as more compact neighborhoods, or defer too much to their wishes. They can pay more for less, because they’re both low density and low population and refuse to change that.
Sammamish has a perfectly fine transit street down the middle of it, 228th Ave SE. It has some ST service now, and ST/Metro plan to increase all-day service in the East Link restructure (2025) and in Metro’s long-range plan (date and funding unknown). I haven’t been to Sammamish except once driving through it, so I don’t know how easily most residents can get to a 228th bus stop, but transit can at least serve the densest third of the city, and that’s something.
https://cdn.kingcounty.gov/-/media/king-county/depts/metro/documents/projects/east-link-connections/routes/269.pdf?rev=223b3806f3f44c42877a77718814c489&hash=C20E92CC805A9F1940390DB16B8845EE
Thanks Eddie. Is this a planned change and part of the eastside transit restructure for when East Link fully opens? I haven’t heard anything about this reroute.
Why would someone take a bus from Redmond to Mercer Island when they can take East Link? Why would anyone from Issaquah today take this bus when they can take the 554 or other peak routes directly to Seattle?
This looks like one very long route which worries me could include a lot of traffic delays. Do you know the total time of the trip? Frequency isn’t that great to begin with.
It is like two distinct routes and riders: Redmond to Issaquah which needs more service, and Issaquah to Mercer Island which I think will need less service, even when East Link opens across the bridge because of the S. Bellevue Station and its huge park and ride. Metro has been cutting peak routes from Issaquah to Seattle/Mercer Island. Why would someone drive to Eastgate P&R to catch a bus to Mercer Island to catch Link (or a bus) to Seattle when they could just go to S. Bellevue park and ride?
I will also point out something I think. Peak on the eastside really doesn’t begin at 6 am, or end at 9 am. 7 to 10 is closer to true peak these days. 30 minute frequency isn’t great to begin with, especially on such a long route that is sure to have traffic delays, especially during peak hours.
I suppose I can’t complain about more service, but like so many in Issaquah I don’t take transit to Mercer Island or Seattle. I need it to Bellevue.
Issaquah resident: most of the network will change with East Link. The I-90 routes that are oriented to/from downtown Seattle today will be changed (e.g., routes 550, 554, 212, 214, 216, 217, 218, 219). Route 269 would have markets at each end: Redmond and Link at Mercer Island; there are intermediate activity centers; riders will use it for many purposes; yes, few riders will use it end-to-end.
Eddie, I am familiar with the Eastside transit restructure after East Link opens across the bridge. Was the rerouting of the 269 part of phase 3? I don’t remember that. If not what is the basis for the post phase 3 restructure of the 269? Probably money.
Granted the 569 today is a long route, and goes from Issaquah park and rides to Overlake park and ride which is a very long ride I took once. I just don’t understand continuing an already very long run to Eastgate and then truncating on Mercer Island. What is the point when Metro is cutting peak service between Issaquah and Mercer Island/Seattle on more direct routes with much better schedule adherence.
I would rather Metro keep the 269 as is and return direct Issaquah to Mercer Island/Seattle peak service. What is the point of 30 minute frequencies between Issaquah — Eastgate — Issaquah?
When you state few riders will use it end to end I agree, so the issue becomes how long the run is. The 569 travels over congested street roads. When you state the 569 has markets at each end I agree about Redmond but not Mercer Island. What activity center is on Mercer Island. Pine Lake is hardly an attraction. I live in Issaquah. I never go to Mercer Island. Who would take the 569 from Redmond or Issaquah to Mercer Island and why?
Is someone in Redmond really going to take the 569 to Mercer Island to catch East Link to Seattle? I don’t see many on the plateau doing the same in large part because the bus trip from Issaquah to the plateau during peak is painfully slow and then you have the walk from bus stop home. They all drive to Issaquah park and rides.
Better frequency from Issaquah to East Link I agree on. Ideally to S. Bellevue, but if it has to be to Mercer Island then I would prefer a dedicated bus from Issaquah to Mercer Island with better frequency and schedule reliability.
Anyway just my thoughts from someone who rides transit from Issaquah, but less today.
In the 1990’s, there was talk of something called “guaranteed ride home” in some metros if you missed your bus. It was basically a taxi coupon. You got a certain small anllocation each year to use.
Of course, cell phone dispatching didn’t exist then. It wasn’t possible to text a household member for a ride either; there were lots of pay phones at transit centers — and in some places the pay phone areas became drug deal call centers.
To me, the role of microtransit is very niche for places where residents are too poor or have bad credit to have a smart phone or hail something like Uber, and the terrain makes walking a half-mile very difficult. I even wonder if it would be more cost effective to hand low income qualifiers a cell phone with limited functionality so that they can see bus arrivals, hail replacement transportation or contact a household member for a ride.
The proper role for microtransit is areas that are too far to walk to the nearest bus stop, or trips that go across the route axis in a way that makes a regular bus trip unreasonably long, and the agency can’t/won’t add a fixed route to address the gap. But there are probably a number of people who ride it almost from bus stop to bus stop, thus abusing the system. They’d do that to avoid riding with strangers or wait for an infrequent route. The largest concentration of trips is in that direction, so it wouldn’t be surprising if people do.
I feel that such programs should not be subsidized out of a transit agency’s budget. These situations occur because of bad street layout, low density zoning limits and other land use decisions made by governmental entities outside of transit operations. If a city wants everyone living on hillsides with low density, that city should foot the bill for providing public transportation and not the countywide transit agency through its constrained budget.
Both PT Runner and Shuttle are in the $50-80 per boarding range. I can get using it for parts of unincorporated, rural Pierce. I do not get them trading a bus for service to Ruston. There is no way that scales, and a bus along the water would be a huge benefit to Point Ruston, which is having serious parking constraint issues, and decrease congestion along a road that would be idyllic for walking and biking, except it a load, cruising parking lot during the summer.
Just ban cars and put in a trolley or something. Runner is not the answer.
The real life example was innisfill
https://www.orilliamatters.com/local-news/uber-remains-in-the-drivers-seat-in-innisfil-despite-some-bumps-7050776
They replaced buses with Uber rides. The rides cost around 4 dollars (note it’s not a fixed price but subsidized) for a year around 95 thousand rides with around 800k subsidy or around 10 dollars per ride.
In order to limit demand and costs it’s limited to around 30 trips per month. And also geofenced to the city.
I think it’s a fine compromise for low density areas. Though for Seattle I guess the subsidy might be a lot higher per trip that I’m not sure if the financial math makes as much sense when scaled up.
Perhaps there could be some middle ground but I do have a hard time writing up how it’d work in detail besides ending up an Uber subsidy
This is a poorly named report. It’s a good bet the union would be fine with Microtransit if the drivers were members. But that aside…
Microtransit isn’t for every market, and like anything it can be done well or it can be done poorly. It can be done at reasonable cost, or high cost. It really depends on the service characteristics and the market being served.
The key principles should be —
to serve and augment the performance of the overall network, ie, attract people to and from the main services, and
to address gaps in network coverage, and
to genuinely enhance public mobility
The advantages this service has include lower labor cost overall, as drivers don’t require a CDL, lower vehicle capital and operating cost, and passenger convenience. For example a 40’ bus with a CDL is a really expensive way to serve a handful of trips that could be served with a minivan driven by a non-CDL operator. These attributes can be leveraged to meet the principles above.
In Lynnwood, for example, it’s working pretty well. The CT Microtransit area serves gaps in their network and connects to key hubs. But not every market has a rail hub and multiple economic nodes.
There is some good data in the link below, tho you have to scroll down and it’s a bit old. I think agencies go wrong with Microtransit when they either try to do too much with it and see it as a panacea, or don’t pay close enough attention to calibrating it to boost network connectivity.
The last thing I’d observe is there’s a legit equity consideration here. It is ok for public resources to be used to enhance the mobility choices for economically disadvantaged people. A subsidized Microtransit trip vs a market rate Uber ride that connects to the bus/rail network is a net public benefit if it helps a person or a family access employment and services they might not otherwise be able to reach.
https://www.communitytransit.org/docs/default-source/board-documents/2023-presos/08-03-23-board-presentation.pdf
The advantages this service has include lower labor cost overall, as drivers don’t require a CDL, lower vehicle capital and operating cost, and passenger convenience.
Sorry, no. First of all vehicle capital and operating costs would be similar (a bus is a bus; a van is a van). The cost *per rider* is much higher. These can be convenient in that they drop people off by their door, but that can also lead to shortages. You get off work, take the bus to the Kenmore Park and Ride and then wait two hours for the shuttle. Yes, eventually it will get you to your home (after going way of the way to drop off other passengers) but that is not exactly convenient.
The only cost saving potential comes from hiring non-union riders. But some of that is skimmed off the companies that operate it. It also doesn’t apply when there is a labor shortage. Transit agencies all over the country (all over the world, really) struggle hiring drivers right now despite the good union contracts. That is because private (non-union) companies will pay drivers a lot of money as well. Of course this includes drivers of the companies that operate these services.
Plus the labor savings are not sustainable. If microtransit would grow then eventually the unions would push back and demand union drivers as well.
Jarrett Walker’s flowchart really sums it up:
https://humantransit.org/2018/02/is-microtransit-a-sensible-transit-investment.html/mircrotransit-flow-chart-02092018#lightbox/0/
If we put money into microtransit instead of fixed routes we end with basically two different outcomes. One is that we charge a lot more money for it. The other is that it is a nicer experience for some, but worse service for many.
In some cases this may be justified. It is also nothing new. Access vans have been in existence for decades now. Hopelink ( a social services nonprofit) has operated this sort of thing for a very long time (https://www.hopelink.org/programs/transportation/). There is a range of services, including DART (Dial-A-Ride Transit) which is run by Metro. DART is largely fixed route, but with advanced notice it can deviate to serve an area. This is an important lifeline for those who are mobility impaired. Hopelink offers other services that are more like micromobility (completely on-demand) but my guess is the systems that are more fixed route help a lot more people (for the amount of money they spend). You just can’t get away from the efficiency of a fixed-route, fixed-schedule system.
I think that is the fundamental problem with micromobility. Folks think of transit as an information problem. I can see why. An all-knowing entity would realize that I’m going from Northgate to Fremont via Wallingford, and coordinate the trip with people going from Northgate to Wallingford or Wallingford to Fremont. But the more you increase service the more the system becomes like a fixed-route, fixed-schedule system.
You can play this out in your head when thinking about an airporter (van). I don’t think anyone who has ever had a long ride from the airport in one of these things has ever thought “I wish all transit was like this”.
But think about it from the companies perspective. You want to get a fair number of riders, otherwise they are no cheaper than a cab. You don’t want the van to go north and then south, otherwise the trip is terrible and the driver is spending way too much time delivering people. Ideally you want to group trips by area or corridor (thus serving people along the way). It is great to drop them off at their house, but this costs a lot of money and is an inconvenience for others. A one-seat ride is great but you soon realize that if you ask people to transfer it becomes a lot cheaper to get people to where they want to go. Basically you’ve just reinvented fixed route transit. (Note: Airporters obviously just pass on the cost of deviating to your doorstep and not having transfers.)
Then there is the timing. Same story. When does the van leave? When the first person arrives? That is expensive, and if you don’t have enough vans it means others are waiting a long time. Maybe when the van gets full, or half full? OK, but again that could mean a long wait. So you run at a fixed time, which minimizes the waiting. In other words, the most cost effective system is fixed-route, fixed-schedule system.
It is easy to think that micromobility is something dramatically new, but it really isn’t. It is simply a refinement on high-cost (per rider) systems that have existed for a very long time.
“You can play this out in your head when thinking about an airporter (van). I don’t think anyone who has ever had a long ride from the airport in one of these things has ever thought “I wish all transit was like this”.”
This comment reminds of a time I used to ride airport shuttle vans from Houston Intercontinental Airport to my parents’ house, 27 miles away. The first time I did it, I waited 30-45 minutes for the van to leave, but I was lucky and was the first passenger to be dropped off, and trip took about 35 minutes. The time second time had similar wait time, but this time, I had to sit through 2 1/2 hours of zig-zagging back and forth before finally being dropped. $40 down the drain for a 2.5 hour slog, and I couldn’t even get out of the van when it stopped at a hotel to get water or use the bathroom without holding everybody else up.
The third time, I decided to not book the shuttle van at all, but just use the regular fixed-route public transit system. The bus from the airport only ran every 30 minutes, but it had a schedule, so at least I knew when it was going to come. When the bus got downtown, I needed to transfer to the light rail, then again, to another bus. But, the light rail was every 6 minutes and the final bus, every 15 minutes. Plus, if I needed to pee, I could actually use a bathroom at any of the connection points without having to feel guilty about delaying other passengers. The total door-to-door travel time, for a fare of just $1.25, ended up being 1 hour 45 minutes, a whopping 45 minutes faster than the shuttle van for about 1/30th the price.
We all love to complain about transfers, but the reality is, frequent service that runs published routes on published schedules works.
That’s like when I went to Newark airport. I arrived in the early evening and was in a hurry to get to the place I was staying in Manhattan, so I took the Amtrak regional and it cost maybe $20. When I went back home, I took PATH to downtown Newark and the airport bus route from there, for a total cost of maybe $3.
A second time, I was with some friends at a retreat in Scranton, and we must have driven back to New York and stayed at a hotel in Jersey City. My flight home was early in the morning, so I left at 1:30am but there was still a 24-hour bus to the PATH station, and I took that to Newark, and the airport bus was also 24 hours, so I got there for around $3 again.
There were actually two bus routes on the street to the PATH station. One was a regular city bus; the other was a PATH feeder bus and it was free. So I took whichever one came first.
This reply pretty much misses the point. Microtransit is not a replacement for fixed route service, nor should it ever be considered that. It can however, be a solid supplement when agencies can afford it, and it can also be strong tool for addressing equity issues. Walker knows a lot and is a very smart man, but he’s arguing a different point, and I disagree with his conclusion on equity.
First of all, on fixed cost: minivan + non-CDL driver < diesel bus + CDL driver. If both vehicles are carrying only a handful of riders per hour, the advantage is clearly with the lower cost solution. Which is to say, some markets — ie some gaps in network coverage — are not always served best with a 40' bus and a CDL driver.
Second, some potential customers have constraints on their ability to reach a fixed route. Maybe they live outside the paratransit zone. Maybe their mobility need is more urgent than paratransit scheduling tools can serve. Or maybe they're lower income people who live in a transit desert for whom a very long walk isn't a good option. Maybe the fixed route up a very long steep hill. These are examples of equity needs. If on-demand service can connect such customers to the fixed route, that is a net public good and a worthy expenditure. It is very much OK to spend public money addressing such needs.
Third, as we all know there are many folks who will never even try transit because it's inconvenient. If on-demand transit can add users to the system by lowering barriers, that to can be value-added. In the private sector, such service might be termed a loss leader, where an investment may not return more revenue, but is made for the purpose of adding market share. If an on-demand service can warm up new customers to using transit regularly, then it's worth considering.
Not every agency can afford to make investments like this. KCM and CT have the benefit of robust sales tax revenue and both leverage significant state an federal grants. so filling in gaps in their networks in this way is appropriate.
We really should focusing on how the service gets subsidized and why.
Is the need is to provide service at a time of day or week, or is it to cover an area that was developed because land is cheap and it costs more per person to serve, or maybe laid out so people can’t easily walk to a fixed-route stop? Is the service to address a promised some larger objective about promised levels of transit service, or is it because someone complained loudly about not having fixed-route service?
I think it’s terrible how we end up dedicating our scarce transit dollars to address situations that transit didn’t create. The larger public arena keeps using transit dollars to fund other big objectives — from a bicycle track on Broadway to replaced utilities under Madison or First Ave in Downtown to this microtransit because some planning department in the 1980’s let someone create lots on cul-de-sacs making them hard to serve by transit.
Maybe we should allow special “micromobility transit districts” that can create a tax and/or fee structure to provide an enhanced quality of service. I hate to see fixed routes stop running in the evenings in a dense urban area with lots of demand to spend the money instead on serving a low density area with micromobility when the service gets less than 1-4 riders an hour.
“…a handful of trips that could be served with a minivan driven by a non-CDL operator.”
This reminds me of the Jitney transports in San Francisco in the 80s and 90s. Private vehicle operators who collected commuters at the major transit hubs and delivered them en masse to various locations across the city, and then reversed course at the end of the day. Fares, as such, were collected by the driver, maybe 2 or 3 dollars. These Jitneys trundled around the city, doing one thing well at a good price point, and were very popular. Can’t have that! Safety! Union drivers! Congestion issues! Smog! And so jitneys, usually small family businesses run by immigrants, were shut out by the deeply progressive San Francisco leaders. These are the same people who go after the Tamale Ladies who make food at home and sell to commuters, because they don’t have food prep licensing.
Let the market solve the last mile issue, it’s the only way to actually get people where they’re going when they need to go.
I’ve wondered if it would be more productive to have a hybrid microtransit/ fixed route operation with “smart transit stops” scattered in the service area that have onsite hailing capabilities. Let me paint a scenario:
The service area has designated “smart transit stops” to get on or off a bus shuttle that aren’t linear but are instead scattered every few blocks so that they are within the asking distance of most or all residences and destinations . Assume that one service area has 15 such stops across an area 1-4 square miles in size. Each stop has a number code. Each stop has a smart pad where someone would hail a shuttle bus or a rider could scan a code at the stop on a smart phone app (so the system know that a rider is waiting) or the user could just hail the service as they left their house. These stops could have small shelters and security cams and lighting activated when there is someone waiting.
Each service area would also have a main layover or hub like next to a Link station. The service would be expected to leave the hub on a designated schedule, like maybe every 20 minutes and be expected to return by that time. (At high demand times the agency could have a shuttle bus leave every 10 minutes so one is circulating while the other one is loading at the hub.)
At the designated time, the vehicle would leave the hub on an instantly determined routing based on two things — where routes are automatically hailed, and where people who boarded the waiting bus at the hub have coded in as their destination stop. With hopefully no more than five hails per 20-minute loop the dispatching and estimated arrival dispatching could be automated.
Something like this could seemingly get more riders per hour than the current shuttle programs can get. However a rider may have to wait slightly longer for a shuttle bus like this. It also would be complicated if a rider was making a trip between two non-hub stops in the area — so that the rider may have to go to the hub first and then use the following loop to ultimately reach the nearby destination.
One benefit is that it would offer onsite use without each rider having a smart phone app.
Its viability may not be fully realized until there are slow-speed autonomous shuttle systems (of which there are well over a dozen companies with them already being field tested around the world). If automated, the agency could simply move around vehicles as needed based on overall demand and anticipated higher special event use.
In a denser and flatter setting like much of central Seattle such a system almost certainly wouldn’t be practical. However it might be well-suited for less dense suburbs located in hilly terrain. It could be more useful in a place like Federal Way or Sammamish for example, where riders often must wait up to an hour of a bus while productivity is low and service hours are limited.
Generally, the idea of smart shuttle service is still a bit in the experimental stage — and technology will keep improving as programs will keep getting measured for effectiveness. So while this example may been too abstract today, some version of it seems inevitable.
The hybrid configuration you’re talking already exists under the label DART. For example, King County Metro route 204 runs a fixed route in Mercer Island, but also has as on-demand service area at the south end of island where people can request (via phone or app) the bus to swing by them before or after the regular service route.
If a stop has an electronic “smart pad” for entering a code, it can just as easily have a single button that sends a pickup request including the location. It would just need a radio transmitter to send the preprogrammed stop ID, and those are becoming every more inexpensive and ubiquidous.
Metro has a precedent of semi-fixed routes like the 904 and Snoqualmie Valley Shuttle I mentioned. These have a fixed route and schedule but they also have dial-a-ride neighborhoods they can detour into. The current ones are hourly or less, and the flexible areas are only a small part of the route, but you can build up from there into anything you want.
I’m not so sure about flex stops around the main route line but not on it, but you can make a case for it. They’re kind of like docking stations for shared bikes. Since efficiency requires stopping at a transit-convenient location on/near your block, these stops could specify where those are in advance.
One type of “flex stop” service that I do think should be a no brainier is rural bus stops that allow passengers to flag the bus down anywhere along the road, even where there’s no bus stops.
Obviously, this does model does not work well in cities, but in rural areas like Vashon Island, it works very well. Provides better service for passengers at essentially no travel time cost. Of course, the bus still has to stay on its route – it won’t detour down neighborhood streets. But it does at least avoid the need for passengers to have to walk along the main road that the bus runs on to reach a bus stop.
Bellevue has the Bellhop. It’s a joke, every time I try to use it it’s full and not accepting more rides… Way too little capacity, way too difficult to hail (using an app), too much coverage area, too much wait. Making it free does not help with meeting the demand and recognizing the limited supply either.
Why is it so hard to just use a downtown fixed-route circulator with clear stops running on a frequent schedule with a large vehicle instead of reinvent the wheel with some worthless transport gimmick (like the Hyperloop, Boring Co., Mobile Lounges, PRT, etc.)?
And that’s the problem with micro transit in a nutshell. When service demand increases beyond a very tiny level, any more passengers cannot be accommodated without finding a bunch of new money to buy new vehicles and hire new drivers. With conventional transit service, a bus has enough capacity to absorb large amounts of latent demand at essentially no cost. Eventually, more buses may still be needed to avoid overcrowding, but the passenger volume needed before that happens is way higher.
Yes. Micromobility services don’t scale. They only make sense at a very low ridership-per-service-hour level. Anything more than that and you are better off running a bus. That doesn’t mean they don’t have value, but it means they will serve very few people for the money you spend (if you are doing it right). They are on the extreme “coverage” end of the coverage-ridership spectrum.
One use is to gather “real data” with regards to an areas potential transit use. If lots of people are using microtransit, they will probably use a regular bus. In that sense it is part of another spectrum, which is based on amount of use. Agencies often go bus, BRT, tram, metro. In other words they run a bus and if it has a lot of riders than they convert it to BRT. If that is crowded they replace it with a tram. If that is full as well, they build a metro. Agencies often skip a step and there are plenty of hybrid systems (it gets a lot more complicated than that) but that is the general flow. Microtransit could be considered the first step (before the area gets a bus).
I don’t understand your Bellhop complaint. Buses and light rail cover much of Bellhop’s Bellevue service area.
On the topic of “why not just subsidize Uber instead”, I think there is value in subsidizing the *availability* of a taxi service in semi-rural areas where having drivers available might otherwise not be economical. Simply having the service available for unexpected contingencies provides value even for people who have cars (what if the car won’t start, for instance), and paying a couple of drivers $10/hour to hang out and wait for passengers all night doesn’t cost that much.
Subsidizing the fares, however, gets really expensive, really fast. And it’s a system where the more people take advantage of it, the more it costs. Even people who can’t afford a car, it’s probably cheaper for the taxpayer to subsidize their rent in a place where conventional transit can reasonably serve rather than pay for a taxi to haul them around every single time they decide to leave their home.
Also, whatever efficiency is gained by not having union workers or the beurocracy of a public agency just gets squandered by Uber’s profit margins. It’s gotten really bad, to the point where at least half of every trip fare is now going to Uber’s shareholders rather than the driver. That’s why even when customers feel like they’re getting ripped off, the drivers are still getting paid crap.
It would make sense to subsidize traditional taxicabs, since they don’t have variable pricing. It would also make sense for the cab companies to run more efficiently by picking up several riders, not just one. Then instead of paying back the agency for the actual fares you pay them for a certain level of service (e. g. three vans covering a particular area). If they pick up a bunch of people or only one, you pay them the same.
That is basically what they’ve done. It explains why it takes a very long time to get a ride in some cases, and is very quick in others.
This generally is a good idea, until you realize that Uber / Lyft drivers will game the system, and since they’re independent contractors its difficult for TNCs to actually direct them on what to do. I saw glimmers of Lyft struggling with incentives that were designed to have drivers where and when they needed them to be in ways that couldn’t get hacked by drivers.
Back when I was a Lyft driver, before I graduated to the XT40, XDE60, and a few others, I had a woman who lived in Sammamish complain that she could never get a ride in the morning. Thing is, as a TNC driver I’m not willing to go off and hang out in an area without some sort of guarantee. I spent far too much time driving places just to wait and not get a ride.
TBH, the solution to her problem would be simply to pay the driver to come from Redmond, Bellevue or Issaquah by making a location there the pickup, having a stop where she actually is, then telling the driver just to say that she was “picked up” at the given location, then wait for the driver to arrive at the actual pickup.
That being said, I tend to think of the taxi/Lyft/Uber business as being a three legged stool:
1. Profits for the company.
2. A livable wage for the driver.
3. Very quick pickup times for passengers.
but you can only pick two. Currently in Seattle 1 & 3 are the legs that are winning.
I found the article to be incomplete. Yes, microtransit doesn’t scale. But other than a token mention, the paper doesn’t address the last mile problem. Improvements to fixed route service don’t help if you have to walk 10-20 minutes on hills to get to a bus stop.
Larry, you’re right. Almost nobody in that circumstance is using transit, and probably never will. Choices have consequences, and the choice to live a mile or mile and a half up some winding rural road with few houses has the consequence that you won’t have transit service and will be completely dependent on the operability of a privately-owned vehicle. Most Uber or Lyft drivers won’t choose to serve you either except in low demand times.
Choices have consequences.
Tom, I agree that if someone lives in a rural area or has a singular driveway a mile or longer they should not expect access to walkable public transit, and I suspect those folks don’t use public transit.
But most experts consider 1/4 mile or 440 yards which is a pretty long distance as the limit someone will or can walk to transit, on one end not both.
There are plenty of urban/suburban/ and semi rural areas that don’t have walkable public transit within 1/4 mile despite many more potential riders compared to your example. The rule isn’t you must live in an apartment in a dense urban area to use transit considering everyone is paying for public transit. Public transit shouldn’t be a tool to force living styles on people. . I live in Issaquah within the ST taxing district and I can’t walk to transit.
In my opinion first mile access is the number one issue that determines ridership, along with wealth. Frequency and schedule reliance are important but you have to be able to get to access.
In rural and suburban areas transit agencies use park and rides because although costly they can serve urban/suburban/rural riders and can get elderly or disabled riders very close to transit without having to wait for Microtransit or make a lot of stops on the way.
All micro transit is trying to do is fill in the gaps or replace park and rides. I don’t think many transit agencies, whether airports, trains or local transit take the position that if you can’t walk to it tough luck.
So far the parking garage or park and ride seems to be the favored way to serve those who can’t walk to an airport, train station, or bus stop although Uber is becoming more popular and will pick you up in a rural area if you schedule ahead but it is usually less expensive to drive and park. Some like airports can charge for parking help offset the costs. Right now I don’t know that there is the demand for Metro or ST to charge for parking despite the enormous sums they have spent on park and rides rural and suburban people seem to like.
The P&Rs are already there. Microtransit is so that people don’t have to use them, or can go from their house to the nearest neighborhood shopping center.
Quarter of a mile is nothing for most people, even half a mile if the walk is safe and pleasant. You walk at least that each way getting from your car to whatever you had to buy at Walmart or Target. The issue is more likely that it can be dangerous to walk even that quarter of a mile, and/or the distance is often much longer than “as the bird flies” due to bus stops with missing crosswalks. Yes, sometimes you are just too far away to walk, and you probably knew that when you moved in. But in many cases it is more of a fixable infrastructure issue.
Of course, if walking any significant distance is an issue, you probably need to be using paratransit anyway.
Setting aside money for the moment, I’m curious how the carbon footprint of microtransit compares with other forms of transportation. Intuitively, I would expect it to rank worse than fixed route buses, private cars, and even Uber. This is especially true if the microtransit is operated with gas powered vehicles, which are especially inefficient for the type of slow speed driving with lots of starts, stops, and idling that is especially common with microtransit.
One problem with subsidizing Uber/Lyft instead is that drivers can and do decline and cancel rides. If it’s out of the way and they couldn’t get their next pickup without deadheading back–decline/cancel. If the pickup is out of the way for a short ride–decline/cancel. Or at least 1 star so they don’t match you again. These companies have cut the drivers’ portion of the fare so much that most drivers have little choice but to cherry pick rides to be profitable. See also: Low star rating–decline/cancel. Didn’t tip? Might get you a 1 star rating.
To confirm a 587 point, the LA and Houston agencies use in-house operators. Here, Metro is using workers from the gig economy.
Per Walker, micro transit can provide coverage in areas that are too sparse for fixed route service. But note that is not where Metro has applied Ride2, Via, Pingo, and now Metro Flex. It has been applied to areas that already have fixed route service (e.g., West Seattle, Eastgate, SE Seattle, Skyway, Tukwila, Renton, Sammamish). Only the Issaquah hills and Juanita lack fixed route service (Juanita Route 935 was deleted in fall 2014). This competition or duplication has to degrade network efficiency. It was not until March 2023 and the Metro Flex branding that the back room computing power was applied to deny trips to those for whom the bus routes were actually better. The Lynnwood and East Link Metro Flex applications are largely duplicative.
One promising avenue is the application of the backroom computing power for Access delivery.
There are two opportunity costs to consider: scarce service subsidy and the allocation of planner time.
Is new coverage the best margin on which to invest? Could weak parts of the fixed route network be replaced by flexible service? Where? There are large parts of South King County without transit service.
Metro Flex does not a fare box. Intending riders must have ORCA or a credit card.
Per RossB and Walker, flexible service cannot scale well; it uses small vehicles; deviation is a feature, not a bug, and those boarding first are dragged through subsequent deviations, slowing their trips and making them less reliable.
Both fixed route and flexible service involve a combination of walking, waiting, and in-vehicle time for riders. What is the net?