After a topsy-turvy first couple of months—first vacations, then snow—traffic on the newly-tolled SR-520 bridge across Lake Washington seems to have stabilized at about 67,000 cars per day. That’s down by about a third from last fall’s level, just before the tolls were first levied.
But data from the Washington State Department of Transportation shows most of the trips that “disappeared” from SR-520 simply migrated to other roads. Traffic on SR-522 around the north end of Lake Washington has increased by about 3,400 cars per day, an 8 percent uptick. And traffic on I-90 has grown by about 15,400 cars per day, or about 11 percent. So of the 34,000 weekday car trips that have “vanished” from SR-520, somewhere between 18,000 and 19,000 are simply taking alternative routes. As we said last year, high rates of toll avoidance were completely predictable; and folks tracking other tolling projects in the Northwest, including the Columbia River Crossing and the Alaskan Way Viaduct tunnel, should take careful note of SR-520’s diversion rates.
Combining all 3 routes—SR-520, I-90 and SR-522—cross-lake traffic has declined by 6 percent, give or take. Of course, vehicle miles traveled for cross-lake trips have likely declined by less than 6 percent, since some folks are undoubtedly taking longer cross-lake trips to avoid the tolls. We won’t know for a while whether tolls have actually taken a bite out of vehicle travel. But regardless, WSDOT’s tolling experiment has proved one thing: tolling even one segment of a road network can substantially alter driving patterns.
Looking through the state’s figures, there are a couple of additional data points to keep an eye on…
The first is the question of how tolls have affected traffic at different times of day. A mid-February comparison of post-tolling 520 traffic volumes (the black lines) with pre-tolling levels (the gray bands) shows that peak rush-hour traffic is pretty close to what it was before tolling.
In particular, folks commuting from Seattle to the Eastside and back have proven particularly willing to pay tolls: traffic eastbound during the morning rush hour, and westbound during the afternoon rush hour, is in the same range as pre-tolling levels. But mid-day traffic is well below pre-tolling levels. Without looking at the actual numbers, it’s hard to know how much of that mid-day traffic is simply diverting to other routes. But if my eyes don’t deceive me, WSDOT’s charts (see page 4 of this pdf, for example) suggest there’s a lot of mid-day traffic on SR-520 that has simply disappeared—that is, SR-520 has gone down more than traffic on 522 and I-90 has gone up. You should feel free to insert your own interpretation here: either mid-day tolls are “too high” because they’re discouraging travel without actually relieving congestion; or they’re “just fine” because they’re helping to raise revenue for the new bridge by charging the people who actually use it.
A second number to keep an eye on is the baseline against which SR-520’s “declines” are being measured.
To gauge the effects of tolling, WSDOT has been comparing the weekly traffic volumes on SR-520 with a recent “baseline”: six holiday-free weeks last October, November, and December. During those weeks, SR-520 carried about 101,000 vehicles per workday. But WSDOT’s most recent Ramp and Roadways Report found that traffic on SR-520 topped 112,000 vehicles on the average weekday—a figure that’s about 11 percent higher than the baseline that WSDOT is using to gauge the effects of tolling. To confuse things further, for years WSDOT has been saying things like this: “Over 115,000 drivers now use the 520 bridge every day.” (Nota bene: This is false: it applies only to weekdays, not “every” day; and besides, weekday traffic hasn’t been that high since 2000!) In short, the “baseline” that WSDOT is using to gauge post-tolling traffic trends is significantly lower than the “baseline” that WSDOT has been reporting for other purposes.
For now, the lower baseline reported by WSDOT is reasonable enough, since late fall and winter tend to have less traffic than the summertime. But as summer approaches, keep an eye on that baseline!! If WSDOT doesn’t adjust the SR-520 traffic baseline upwards to account for the summertime boost in traffic, I think it’d be fair to say that they’d moved the goal posts.
Cascadian
Here’s something that jumped out at me: there’s a pretty large discrepancy between eastbound and westbound traffic, most of it showing up on 522. 6600 more cars head west compared to those that head east. That’s 26% of the westbound traffic returning by some other means each day, unless we want to postulate thousands of cars abandoned along westbound 522 every day, which I think I’d notice. Presumably they’re using surface streets, or going north via the I-5/405 interchange, and a small handful really are one-way trips. But why would people be so much more likely to take 522 westbound and take alternate routes back?
Clark Williams-Derry
Wow, clever of you to spot this. I have no explanation for this. By my read, I-90 has roughly the same amount of eastbound and westbound traffic, but both 522 and 520 have more westbound traffic than eastbound.
The only reasonable explanations I can think of are: (1) faulty measurement, and (2) weird traffic patterns, e.g., people commuting from Seattle to the northeast of Lake Washington via I-5, to avoid both tolls and backups on 522. But seriously, your guess is as good as mine.
Sebastian Helm
The pattern on 522 could be explained by people having to commute on time for the beginning of their work day, but having more flexibility or leisure (for detours, shopping, or other) in the evening. But what about the behavior on 520 itself – why are eastbound commuters (EB in the morning, WB in the evening) so much more steadfast than westbound ones? Could part of this be explained by public transportation? Are the numbers for that in, yet?
David Emerson Feit
Notwithstanding measurement error, I think this clever observation illustrates that this set of roads doesn’t constitute a closed loop. The traffic must be going somewhere. If we were imagining this as intra-metro commuting only, then it really would be surprising to find weird alternative routes where the eastbound traffic is diverting. But if we extend our imagination to trucking and tourism, then it’s easier to believe we’d get seemingly asymmetric patterns that actually resolve to better symmetry beyond this set of roads.
Clark Williams-Derry
OK, I couldn’t help myself. I took a look at the numbers a little more carefully, and decided that I could divide trips into “northish-eastish” vs. “southish-westish”. The totals for the two categories together should be pretty close, though maybe not 100% on because some roads aren’t counted. (SR-99 and various surface streets and smaller roads, for example.)
The decision to combine northish with eastish was a bit arbitrary, but it made sense to me given the geometry of the road network. I-5 comes north out of Seattle, but then curves a bit to the east…making “northbound” I-5 serve some of the same destinations as “eastbound” SR-522, for example.
When I lump those categories together, the “northish-eastish” and “southish-westish” basically balance out — there’s a difference of less than a percent. The difference is probably within measurement error, but can also be explained by roads that aren’t included in the data, e.g., SR-99. Some people do drive I-5 south and take SR-99 back north, or vice versa. (I’m occasionally one of them…)
So maybe there’s no mystery here, just a curiosity of the road network and the particulars of Seattle’s geography.
Danner
They have to be at work on time(am) but they don’t have to be home on time. Hence other, slower ways of commuting home.
Morgan Ahouse
Maybe we need to include 405 and I-5 in the analysis, even if we expect those flows to reflect more interior patterns.
Jon Stahl
Is WSDOT actually publishing daily traffic count data anywhere? And if not, why not? Shouldn’t we be advocating for “open data” here so that we can make accurate independent assessments?
Clark Williams-Derry
After a few weeks of publishing daily numbers, they now publish weekly numbers on their website:
http://www.wsdot.wa.gov/Tolling/publications.htm
But it’s not a dump of all of their data; just a summary. So there are details that I can’t figure out without a formal data request (plus some time and patience).
Morgan Ahouse
We’ve talked alot about peak reduction, but I’m thinking that the lack of infill from peak to shoulder shouldn’t be so surprising yet, as people are still adjusting their committments at work, day care and the like. When vehicle and housing leases renew, old vehicles wear out, and the like, I expect we’ll see a gentle and steady drop in these numbers.
My vote is that the midday rate is too high.
Dow WSDOT plan to survey peak travelers? I’m curious about the demographics.
Clark Williams-Derry
For what it’s worth, I can’t find any significant trend on I-5 or I-405 after tolling started. Traffic may be down a wee bit on some segments, but it’s been within the normal range of variation.
The only exception is 405 between I-90 and 520, which is up a bit — which provides more evidence of toll diversion.
Daniel Hornal
We really won’t know much for about a year. People who live or work south of 100th and north of downtown and commute to Bellevue/Redmond/Kirkland have very little choice but to take 520. Over time we might see a lot of people recognize the expense of the tolls and adjust their patterns. Also, the way that tolls are set up, it’s not an “everyday expense” so it makes sense for there to be a lot of lag in pattern adjustment.
Dan
On the flip side, we might see that some people that are currently avoiding the tolls come to realize that it’s too big a hassle to travel out-of-direction and decide to pay the toll on the 520. I’ve read studies of other bridges or roads that have tolls added after they’ve been in place for some time showing similar patterns of avoidance, then acceptance. Wonder if that will happen in this case? Hope you’ll keep an eye on it, Clark. 🙂
Clark Williams-Derry
I’ll definitely keep an eye on the trends — though I’m not sure I’ll have anything interesting to report.
I was actually surprised that traffic patterns & volumes stabilized so quickly. I expected to see steady evolution in the numbers; but barring vacation and snow weeks, they’ve been surprisingly static so far!
John Niles
Does anybody know why WSDOT didn’t include the I-405 route around the south end of the Lake in the flow monitoring? Wouldn’t higher volumes on I-90 in response to SR 520 conditions cause some diversion from I-90 to the southernmost go-around?
Another question, about diversion to transit, a big expectation in the planning for the SR 520 tolling during the Bush/Mineta/Peters “fight gridlock now” administration at USDOT http://web.archive.org/web/20081201181132/http://www.upa.dot.gov/index.htm — I’m hearing that relevant park & ride lots and bus runs for former cross-Lake drivers westbound in the morning are frequently chock full and thus discouraging the transit alternative. Are data yet available on the transit conditions in this new era? I do understand that the full complement of extra buses planned in the Urban Partnership Grant were deployed, but that this incremental addition might now not be enough to avoid excessive SRO on cross-Lake corridor buses.
Final question — about commuters from Seattle to the Eastside in the morning. I’ve heard the contention that the Metro eastbound morning transit service for Seattle outbound is significantly less than Seattle CBD oriented and thus bus crowding may be potentially worse eastbound in the AM and westbound in the PM than the opposite direction. Can anybody confirm or deny, easily? I realize I could study the schedules, but haven’t and maybe somebody else has, or a Metro employee who reads this could report publicly. Thanks!
Clark Williams-Derry
John —
On your first question: They actually did look at 405 around the south end of the lake — but traffic volumes actually haven’t changed much compared with the baseline level.
The other 2 are *great* questions, but I don’t know the answers to them. Anyone else??
John Niles
Maybe WSDOT somehow knew all along that nothing would change on the south part of I-405.
Kathy Paul
Another question not addressed in this discussion – was the stated purpose of the tolls to force people into public transit (in which case, not enough buses were laid on, at least for the a.m. commute)? Or, is the aim of the toll to raise enough funds to build the new bridge? If it’s the latter, then it’s failing – people are driving across (commuting for free, I should say) 1-90 in DROVES, completely clogging 1-5 through downtown Seattle and making the 1-5/1-90 interchange a nightmare for much of the evening. It seems to me that the toll has (a) damaged the flow of traffic across I-90 and is not succeeding in raising the projected $$ for the new bridge. All of which was predicted by anyone who drives across either bridge in the morning or evening… but not, it seems, by the DOT.
Levin
How substantial a drop in traffic would it take, for you to be impressed? Do I understand correctly that 15% (15,000 of 101,000) of cross-lake car trips disappeared because of tolling? In most discussions of transportation alternatives, a 15% drop in car trips is worthy of celebration, rather than a disparaging subtitle. When I-90 gets a toll, maybe that’ll knock off another 15%?
Clark Williams-Derry
When I do the math, it’s just 5 to 7% total, depending on the week. And higher gas prices may be responsible for some of that decline.
If it were 15% I’d be impressed!
K kerl
Tolling was put in too early for alternative transportation to kick in; not having access to the proposed new bridge’s benefits! where is my 520 bike lane?? I’m dropping out of commuting east, and now work out of my Seattle house; that probably accounts for a few thousand of us former dirvers. In the interim a good deed would be if passengers WITH bicycles could ride the cross-lake busses free, reducing city car traffic on both sides!!
Morgan Smith
I don’t commute to Seattle — I live and work on the Eastside — but I used to take regular weekend trips into Seattle using the 520 bridge route, since it was the straightest shot. Since the bridge toll was instituted, I simply don’t do that any more. Seattle has made it clear that people in automobiles are anathema. I’ve responded by spending my small discretionary income on shopping, dining, etc. at Eastside venues or driving north to Lynnwood. Yes, I miss Pike Place Market, but not so much that I’m willing to pay insane “auto penalty fees” (tolls and parking charges) to take in its ambience.
RadWriter
It seems to me that one ought to consider improvements or worsening of the local economy in examining traffic volumes. I know that when I drove I-5 back in 2004 – 2007, traffic was a lot heavier than it was in 2009 and 2010. If the economic improvement is realized by daily commuters, that could increase cross-lake traffic and explain some of the increase relative to an earlier baseline. The time of year may also do that. Are there any good stats on autumn/winter traffic vs. spring/summer?
While this may not be the place for the more general discussion of tolling as a public policy, it is something that we ought to consider. Since one form of revenue merely replaces another, on whose behalf are the tolls imposed and extracted? Are the banks and those who escape other forms of revenue generation the real beneficiaries of tolling?