If traffic statistics are to traffic fans as sports statistics are to sports nuts, we who are addicted to the traffic pages should have no embarrassment indulging our fascination in the SR 520 tolling project daily data feed from the Washington State Department of Transportation: a full-color array of 20 maps, 162 explanatory graphs and 37 tables on traffic waves and ripples all around Seattle and its suburbs from the SR 520 tolling.
Unfortunately, as befits the topic’s true complexity this detail, which has to be requested, is intriguing to gaze upon but both too detailed and too transitory from day to day to neatly summarize. So, like baseball slugging percentages and football quarterback efficiency ratings, a bit of selection and personal opinion is necessary for stimulating spirited discussion.
But first, some context. The start of tolling on SR 520 followed by less than two months the voters’ rejection of Tim Eyman’s I-1125 attempt to slam the door on modern tolling in our state. It’s important for Olympia to have heard unequivocally that voters are prepared to see what may happen with tolling. From my own efforts opposing I-1125, including a lot of opportunity for thought about where tolling might be appropriate and how much money it could raise, I came away with great caution toward bullish suggestions that tolling offers more than a partial contribution to the gaps in future transportation funding.
Tolling must be used as a fine-edged tool, not a financing blunt instrument, because the traffic-altering byproducts of careless tolling could have negative and inefficient effects instead of the positive outcomes we should be seeking. And I continue to strongly support the I-405 Express Toll Lane program, which will eventually add a lane in each direction from downtown Bellevue to Woodinville/Bothell.
As on I-405, tolling’s great promise is something apart from a mere construction-financing tool. This more powerful promise — necessarily using variable pricing that constantly tunes the price to fit the immediate circumstances of traffic demand and road supply — lies in the ability to make roads flow more smoothly and produce efficiency benefits, increasing the economic value of infrastructure investments. (In the I-1125 discussion, many voters and commenters clearly got this basic economic point; I-1125’s attack on variable tolling was one of its fatal paths to downfall.)
The greatest importance of the SR 520 program lies in what it will show about making roads more efficient. It’s why the results are being watched not only here but also around the country, and it’s precisely why the U.S. Department of Transportation in 2007 committed more than $150 million to support this program across Lake Washington.
The most encouraging results in the traffic numbers so far are not surprising, but interesting, problematic and unsettled questions also lurk in the details of the early data. It’s already possible to see the huge potential for tolling to point us toward ways to improve our transportation system. So it’s important from the outset to look at implications in the early numbers that remind us how carefully this path must be taken.
For me, the performance of a modern tolling program is chiefly important in relation to two meta-truths about transportation systems and traffic. The first is that traffic congestion is a product not principally at least of too few lanes of roadway but of too much unevenly timed demand. This is the heavy peak use — the daily morning and afternoon surges of drivers and their vehicles that clog the otherwise largely open roads into stop-and-go molasses.
The second is the counter-intuitive but widely agreed upon fact that during peak periods, a large share of the cars and drivers — mixed in with the expected work-commute trippers — are making non-work related trips. Generally accepted estimates of non-work-related volumes run to 40 to 50 percent of peak period trips. Those trips can’t all be made at another time or in another way, but many of them can or will be when it’s in the drivers' and passengers' interest to do so.
We’ve invested untold billions in building and maintaining highways and we face staggering monetary, environmental, and social costs of undertaking to build even more. It’s therefore axiomatic that gaining more efficient use of what we already have — the roadway stock (including the transit systems that run on the roads themselves) — is the first imperative of transportation improvement policy. That’s why low-cost/high benefit programs to speed the clearing of freeway traffic incidents became so prominent on WSDOT and the Washington State Patrol’s agenda almost a decade ago, when experts around the country were noting that traffic-disrupting incidents play such a big role in many of the worst back-up nightmares.
To tolling: the big payoff for making efficient use of roadway investments would be to reshape the daily curve of traffic demand. You can then shave the peaks, fill in the valleys, smooth the traffic flow across the day, save fuel and reduce emissions, speed freight, improve the reliability of expected travel times, and cut everyone’s personal exposure to delay and aggravation.
And, by improving the efficiency with which roads are used, we also can move to capture the really big savings in future transportation costs and trouble, because we then can gauge and pursue our needs for more investments based on how roads should perform, not on how we suffer today from traffic performance in the daily boa constrictor grip of peak-hour congestion. Efficiently used roads will always be smart infrastructure, helping to support a productive economy.
SR 520 tolling really is a national breakthrough for showing how to manage existing transportation systems as if we really cared about the efficiency of road performance and how to do so with the forward-looking techniques of toll collecting and accounting. However, like many groundbreaking efficiencies in any economic sector, SR 520 tolling is an imperfect or, perhaps better stated, an incomplete experiment.
The main flaws in SR 520 tolling start with the unwieldy design of the tolls to apply only to one part of a single cross-Lake Washington travel corridor (something created by politics). Thus the value of the program is diminished by the distortion of traffic demand bulging onto “free” I-90 rather than reshaping itself within a larger, rational cross-lake pricing plan that would embrace both SR 520 and I-90.
Secondly, the fixed and inflexible pricing tiers in the time-of-day toll schedule adopted by the state Transportation Commission and the Legislature are a weak substitute for truly variable market pricing. This frustrates the full power of off-peak discounting to maximize the use of SR 520’s all-day capacity. Only with flexible, fluctuating market pricing, tuned directly to traffic demand levels on the roads, will tolls achieve their full ability to make roads efficient.
Nonetheless, the SR 520 program is a big stride in the right direction.
What can we see in the early results? And what can we still expect to learn as drivers make their own weightings of time, distance, and money and settle into new patterns with a tolled SR 520 in the traffic management mix? Think of the big diversion question this way: It’s not most importantly about the avoidable downside of diversion to I-90, the byproduct of shortsighted politics. It’s about the effect pricing should have to divert, that is to reshape, SR 520 traffic in different patterns across the day on SR 520 itself. Success will improve the economic and social return on investment represented on SR 520 itself, a major achievement if it happens.
The big news so far is no surprise. The peak period price ($3.50 a crossing at the five hours daily at the height of customary travel) has dramatically shaved peak volumes on SR 520 at the same time it has handed remaining SR 520 drivers the benefit of slashing their time in traffic. The number of drivers who judged the savings in time worth the toll has climbed day after day since tolling began. Last Thursday and Friday mornings’ (Jan. 5 and 6) eastbound traffic volumes on SR 520 were at about 75 percent of pre-toll levels, yet traffic continued to speed along, a transformation of the old SR 520 traffic experience.
By Thursday evening’s peak period, westbound traffic at one point came very near to the typical pre-tolling experience, yet continued to move at good speeds, reflecting another well-documented traffic fact: Very modest easing of traffic volumes can turn clogged highways back to reasonable speeds. .
Less happily, off-peak traffic seemed, on the early experience, to have sustained a very large reduction in volume, dramatically and disproportionately reduced from pre-tolling experience. The off-peak price ($2.25 for one evening and four midday hours) seemed not sufficiently discounted (especially considering that in these hours the alternative on I-90 features no price and no congestion) to draw the traffic to the SR 520 Bridge that it easily should handle.
This suggests a hypothesis that the SR 520 off-peak price should be lower: More people would use the SR 520 bridge, rather than choosing, for many of them, a longer gas-guzzling detour over I-90. With greater volume at a lower price, midday revenue from the SR 520 Bridge could actually be higher. With true variable pricing, this issue would be solving itself.
By late in the first week, however, the midday numbers seemed to be showing a somewhat better trend. Early in the week the noontime traffic volume on SR 520 was well below half of pre-tolling experience and the graphing of volume across the midday hours resembled a deep ravine. But as the week ended, that noontime number was slightly over 50 percent of the pre-toll volume, still too low, but the graph line of volume drawn across midday between the peaks began to resemble a more attractive gentle swale.
Data for the shoulder periods (just before and after the peak hours) is very interesting. On SR 520, the price is lower (at $2.20) before 7 a.m. and after 9 a.m. (and for the afternoon, before 3 p.m. and after 6 p.m.). On I-90, drive times are faster (a lower price in congestion) in the shoulder periods. The evidence is that the shoulders are starting to see proportionally heavier use. That's exactly what should be expected in light of the fact that so much peak period traffic actually is made up of drivers and vehicles not rigidly tied to peak period commuting. But how strong or lasting the trend will be remains to be seen.
All of this early data falls against this important overview. Week one showed not just a shift in route and time travel patterns, but an appreciable drop in overall cross-lake vehicle traffic when SR 520, I-90, and SR 522 north around the lake are combined. WSDOT has put out a typical pre-tolling benchmark of about 280,000 cross-lake trips each day: 101,000 of those on SR 520, 138,000 on I-90, and 41,000 on SR 522. Against the benchmark slice of 101,000 for SR 520, last Tuesday 48,000 of those trips were missing from SR 520, and about 11,000 extra trips showed up on I-90 and SR 522. By Thursday, the meltage against the benchmark on SR 520 was down to 42,000 and the uptick on I-90 and SR 522 was up to about 25,000. So, even as traffic by Thursday was moving towards a new equilibrium, about 6 percent of the total crossing pre-tolling volume estimate couldn’t conveniently be accounted for on the other two routes.
Sorting out that early riddle will bring some of the most important information about the SR 520 program. As time passes, some of the missing cross-lake trips may reappear on the roads. Some may not cross the lake at all: a shopping trip closer to home! Some may move into virtual space: perhaps a worker will shift to a day a week of teleworking.
But an important clue to look for as soon as tomorrow (Jan. 10) is Metro Transit’s first report on how many new bus riders, including on Sound Transit express buses, showed up last week. Park-and-ride lots were said to be filling earlier, and buses were reported to be fuller than usual. New bus riders saved travel time in peak periods from an uncongested SR 520, just like the toll payers, but they did so by paying a bus fare, still a discount from the toll, and many of them they may also have saved parking costs at their destinations.
An appreciable gain in shared-vehicle ridership — not only on Metro and Sound Transit buses but also on private mass transit like the Microsoft Connector — would be a powerful gain for efficiency on the highway system. If the transit statistics reveal signs of that gain, it will underscore what many have long been convinced of: the significant latent power of transit service on SR 520 to make an ever-larger contribution to an efficient regional transportation system.
That neatly sets up a special perspective on one of the largest transportation planning issues now facing the region: What will the final plan look like for completing the SR 520 replacement program on the westside connecting from the lake bridge itself to Seattle and I-5? Will that plan adequately support the most effective long-term future use of transit on that route across the lake? That’s a critical next step in the discussion.
Here’s a clue: It’s not about light rail on SR 520, a terribly suited option for moving large numbers of people among the dispersed residential and job locations east and west that SR 520 quite literally bridges. Rather it’s about designing a complete infrastructure investment for SR 520 and its connection to I-5 and elsewhere that furthers the potential for tomorrow’s modern, flexible transit suited to our economic geography. And it's also about taking advantage, as future transit will, of emerging transportation technology.
A future smart bus likely will look and operate as little like today’s transit coach as a smartphone resembles a hardwired rotary-dialed telephone. Tomorrow’s transit fleet will be far less dominated by rubber-tire buses, although they will still be important. What we will come to think of as transit will also see vast and convenient expansions of powerful commuting tools foreshadowed by the Microsoft Connector and King County’s nationally admired vanpools.
And many of tomorrow’s transit users will first park their electric cars at a park-and-ride lot, where they will choose from an array of conveyances for the rest of their trip to a worksites. These are the springboards for visioning the future power and use of transit on SR 520. SR 520 must be fashioned from end-to-end to anticipate a big role for that coming transit transformation.
Tolling shows we can put SR 520 up to speed. Fitting the rest of the package for SR 520’s future to make the most of that potential is the next big hurdle.
