I was all set to slay Sound Transit for leasing 103 parking spaces in Edmonds for $150 per month each, following a news cycle of recriminations over high operating costs. Park-and-rides are necessary evils, good money after bad, and all that. But when I sat down and worked out the numbers, I found that it may be one of the more cost-effective things an agency has done in a long time.
Let’s build a little model with the following assumptions:
- Every space is full and brings a single rider, or two boardings. No carpools.
- Each boarding generates the 2011 North Sounder Average of $3.30, as stated in the 2011 Fare Revenue Report (p. 24). Edmonds is the lowest fare to Seattle, so that’s probably a slight overestimate.
- None of the lot users previously walked, biked, kiss-and-rode, carpooled, or took the bus to the station because it was too hard to park. We simply don’t have the data to evaluate this properly, though I suspect this is more true than not.
- The marginal cost of carrying a rider on the train is zero.*
Assuming 255 work days per year, the parking contract comes out to $3.55 per boarding, meaning the net subsidy is 25 cents per marginal ride. That’s an order of magnitude lower even than putting on another bus, providing lots of margin for the model assumptions to be wrong while still making it a relative bargain. It’s one advantage of building attractive high-capacity transit: simple schemes can boost ridership and attract high fares for little or no net cost.
* According to ST’s Kimberly Reason, insurance and 1.96% state tax on fares mean this isn’t strictly true.
Regarding the footnote, I thought Norman told us that transit users don’t pay taxes on their fares.
Really — taxes on fares?
Thanks Sherman for working out the math. Let’s hope it works to increase ridership.
I had the same reaction. The state is charging special taxes on railway tickets? Why, for the love of all that is competent? Does the state charge the same taxes on bus tickets?
If this were just part of sales taxes it would just seem to be a matter of “uniformity”, but it obviously isn’t….
This seems like a sound analysis! The only problem I would point out is that adding new free parking schemes here will encourage others to demand them in places where they won’t work, although I’m not sure that should be a primary decision-making criterion.
Suppose that same $3.30 per ride subsidy were instead spent to reduce fares to the level of the 510? Would that attract the same number of new riders for the same cost? Those that wanted to would be free to invest the fare savings to pay for their own parking.
I also question the notion that a full lot will really mean an additional 103 people riding Sounder each day. Some of the spaces will inevitably be used by people carpooling onto the ferry for trips to the peninsula. Others will be inevitably used by employees at the nearby retail stores who want to keep the limited parking in front of their store available for customers.
Unless ST plans to spend a ton of additional money policing its lot, I think it would be fair to say at least 10% of the 103 spaces would be used for non-Sounder-related purposes.
I don’t think lowering fares will significantly impact ridership. Raising fares on Metro hasn’t. For the people who get a benefit from transit, there’s a cost chasm between that and driving.
Gas prices do more to drive ridership than transit fares. At $5/gal a lot of people with long commutes opted to use transit. At $3.25 not nearly as many will trade time for money on a long commute and short commutes the marginal cost of driving ends up being cheaper and more convenient than the bus. It’s really hard for me to spend two hours plus and $5 on Metro when I can spend less than $4 in gas and the commute is under 15 minutes each way.
Ok. So if free parking vs. paid parking means more ridership, why doesn’t free fares vs. paid fares also mean more ridership? At the end of the day, it’s the total cost that matters. How much of it is train fare vs. parking fee shouldn’t matter.
Edmonds to Seattle is 16.8 miles. That’s a 30 mile round trip or about 1 gallon of gas in a modern car…$4. Add in all day parking and it’s a $20 cost round trip…or the price of the unsubsidized Sounder round trip (from a previous article).
Then you’d get fewer riders for any level of subsidy, as you wouldn’t recover as much as you do with the parking.
asdf, the multi-county adult fare for the 510 is $3.50.
The fare from Edmonds to either Seattle or Everett is $3.50.
So, the fare is already at the level of ST Express fares. ST Express is $0.25 cheaper for RRFP fares.
Just when I think I know where STB stands on an issue, a blogger says something that makes me question my assumption. Is STB for park and ride lots or not? At other times, they’ve argued against the idea of people driving from their home to a parking lot to catch transit, encouraging them to walk, bike, or take a bus. Now it seems they are endorsing the driving part of the trip. Can I please get a definitive answer on what STB believes?
Sam, it’s pretty simple. Every author has their own views and writes under their own name. We disagree with each other frequently and sometimes furiously. If you want something that “STB BELIEVES” wait for an unsigned Op-Ed, of which we run only a handful per year.
To add to what Zach says, I’ll repeat something I have said to you before when you have tried this troll technique: Different Situations call for Different Solutions.
Context does exist, and it does matter.
It’s much more complicated than that. You’d be correct if you really got $3+ in full fares, but I’d bet the real fare per ride paid is significantly lower due to employer-paid Passport accounts, which account for a good portion of Sounder North’s ridership. The current rate for Seattle CBD employers is $612/employee/year, or $51 per month, or between $1.10-$1.25 per commute trip. Furthermore, these accounts will already have been paid in full by the employers, and getting new Passport riders (who previously drove) wouldn’t generate ANY new revenue until the next Passport renewal cycle took the higher usage into account when determine contract price increases. If CT-Express riders with Passport accounts switched to Sounder, it would only represent a shift of shared ORCA revenue away from CT to ST.
Yep. I agree – this is a lot more per new rider than $0.25.
Zach,
The $3.30 per boarding figure is ACTUAL revenue, including all employer discounts and such.
My compliments to Martin Duke and his ability to convince me ST isn’t really spending $927,000, they’re really just spending “25 cents per marginal ride.” Yes, 25 cents seems very reasonable and palatable. $927,000, not so much.
I am surprised. North sounder really needs more riders so anything they can do is worth ay least looking at.
Giving away train rides for free will never ever result in an increase in the fare recovery ratio. It’s the lose a little on each one and make it up in volume idea.
There are other benefits to consider. Theoretically, getting 103 extra people to take the train and not drive also takes 103 cars off of the highway heading to work. No? If that can be done, and all we do is break even (or close), then we are ahead.
It’s simple math. Let’s start with free parking, but a $7 round-trip train fare, which is what we have now, and assume that everyone who rides the train drives to the station and parks there.
Now, let’s suppose that Sound Transit inverted the proposal so that parking cost $7 per day, but the train ride was free. In theory, the cost to the agency to operate the trains would be the same while the cost to the riders would also be the same – whether the $7 per day is spent on train fares or parking, $7 is still $7.
Now, starting from that proposal, let’s suppose that instead of ST spending about $7 per space per workday to lease the parking lot from a private owner, only to charge that $7 right back to the train rider, ST just did nothing and let the private landowner charge the customer the $7 for parking directly. It’s exactly the same as before, except this time the private landowner, not ST, absorbs the loss of revenue if all parking spaces aren’t filled up all workdays.
So, whether we spend an additional $7 per rider per day to subsidize the parking or the train ride directly, it really makes no difference to the budget. The National Zoo in Washington D.C., in fact, uses this same principle in its pricing – admission to the zoo, itself, is free, while parking costs $12 per car per visit. (Parking is also free if you buy an annual membership to the zoo).
ST’s leasing the spaces is not really about making more parking available. Those 103 spaces are already available today for anyone who wants to use them – you just have to pay.
What it’s really about is saying that:
1) People who want to drive to the station need a greater subsidy than people who arrive by ferry or walk/bike/bus there.
2) A dollar on spent parking and a dollar spent on train fare, while rationally equal, are not really equal in people’s minds – psychologically, people are ok paying money for a train ride, but parking is expected to be always free and when it isn’t free, they feel ripped off. If this is really true – that a $7 per day parking fee would be equivalent to, say, a $14 round-trip train fare, then subsidizing parking does become much more efficient than subsidizing train fare. But only if people insist on acting irrationally. Better would be if people would simply learn to act rationally and look at the total cost, compared to what their getting and not dwell over how much of it is in this bucket vs. that bucket.
I like your way of looking at it.
My analysis was far less detailed. I saw a price of $150 per month per space, which seems kind of high for a suburban parking space that’s already built and is just sitting there with a “For Rent” sign on it.
And, of course, if ST is going to go into the business of giving away parking, I’m not sure why Edmonds should have all the fun.
I’m not sure why Edmonds should have all the fun.
Because it is served by a line with significant spare capacity, relatively low TOD potential, and relatively high fares.
Martin has estimated the cost of a marginal auto access rider on sounder. the opportunity cost is how many riders could be attracted if the funds and land were used differently; the funds could pay for additional trips on Snohomich County routes 510, 511, or 512. they attract good loads.
eddiew,
I think the data suggest that wouldn’t pay off quite as well. Adding new trips means each rider costs at least $4.05, assuming we match the mean performance of the cheapest route (the 511), which is a pretty tall assumption. In this case utilizing the Sounder capacity is cheaper.
Given that you still have to have extra 510/511 trips ready go whenever there’s a mudslide (enough to carry everyone who would normally take Sounder), it’s unclear to me how much “utilizing Sounder capacity” actually saves anyway.
Or is the working assumption that people will drive to Sounder, see service is canceled, then get back in their cars and drive all the way downtown.