President, Metro Analytics
Finally freeways CAN run well again…
- without being widened again!
- without inducing more demand!
- and with a good shot at political viability!
Urban freeways are a vital part of America’s economy, but at peak hours they collapse due to overwhelming demand. Is there anything that can be done? YES! It is possible to get freeways running again without making them even larger, through a concept called “Managed Motorways.” This is article explains in “Plain English” the fundamentals of freeway congestion, and how to optimize them to perform reliably regardless of the scale of demand. It also explores sustainable enhancements to Managed Motorways so that freeways can always function at high speeds, but not accelerate sprawl. Topics include:
- Freeways, 101 – What Causes “Stop-n-Go” on Freeways?
- Managed Motorways, 101 – a Cure for Stop-n-Go
- Utah’s Citizen-Driven Managed Motorways “Shared Solution”
- How Managed Motorways Can Replace “HOT Lanes”
- How Managed Motorways Can Include Free Market Pricing
- Managed Motorways Will Increase Sprawl and Induced Demand
- Modify “Congestion Pricing” for political feasibility via “Congestion Credits”
- Return on Investment – This is Worth a Try! It’s Worth $Billions of Tries!
- Political Feasibility via Performance Based Planning
- Summary and Conclusion
Freeways, 101 – What Causes “Stop-n-Go” on Freeways?
Love them or hate them, freeways are a permanent part of America’s urban fabric. In past decades they contributed heavily to our economic vitality. Now freeways are crawling at peak periods, and in many cases it’s nearly impossible to make them larger. Engineers often recommend “Big Digs” anyway, perhaps partly for the massive revenues, but also largely because there don’t seem to be better solutions. To understand how to keep freeways working, it helps to appreciate what causes them to fail.
Freeway Failure, 101
Your car’s dashboard may have a “red line” where if the RPMs ever get to that level, you’ll likely ruin the motor within minutes. Transportation engineers are taught that for freeways, that red-line occurs at about 2,200 vehicles per hour per lane. Yes, they will operate at this 100% capacity mark… for about 10-minutes. But as soon as a few more vehicles attempt to jam in it causes a wave reaction. Here’s the common scenario that occurs virtually every day on virtually every urban freeway.
Say it’s 5:05 pm, and you’re moving along well enough at about 55 mph. Vehicles are crammed together and there is no space to change lanes easily. Odds are, each lane on this freeway is carrying about 2,200 vehicles per hour at this point. As you approach an on-ramp, you see a vehicle that needs to merge in, but you can’t change lanes yourself, so you break just a bit to let them in. Because you hit the breaks, the vehicle behind you had to also break. The vehicle behind him decides to change lanes, and that forces those in the next lane to have to break. You get through fast enough, but you can’t see the miles of chaos behind you from the chain reaction that the car in front of you started.
Figure 1 Illustration of how congestion also reduces throughput (graph based loosely on real data, though more research needed to confirm effect).
As more vehicles enter without “gaps” for them to fit into, traffic has no choice but to break and let them in. Within minutes, what started as modest “break tapping” descends into full-fledged, “stop the car or you’ll crash.” Thus when you’re already running at the red-line of 100% throughput, and 102% attempt to fit in, the freeway will break down. Like electrical overload, you just can’t draw 22-amps on a 20-amp circuit without tripping the breaker.
You might say, “Slow is bad, but at least 102% are moving through slowly, right?” Wrong! Where the system was moving 2,200 vphpl, most likely it has now dropped by 30% or more to 1,500 vphpl or lower, just at the moment when more than 2,200 vphpl are trying to get through! Figure 1 is illustrative only and based only loosely on data points I have observed from traffic counters in Utah, and also a few studies in southern California. But it illustrates how as speeds get lower (reflective of stop-n-go), throughput also gets lower.
Figure 2 is another way to look at it. The “laws of physics” mandate that the lanes you have will not carry more than 2,200 vphpl, but that hasn’t stopped 2,500 who arrive at five from trying to force their way in anyway. Ramp meters, as currently deployed in most of the world, slow these extra vehicles from getting on, and hence delay the collapse. But the meters are not smart enough to know when they’re still inadvertently letting too much on, nor are they coordinated with upstream ramps to know how to optimize flow.
Figure 2 When demand exceeds the critical 2,200 vphpl threshold of stable throughput, actual throughput sadly collapses to just 1,500 (70%) or lower. But Managed Motorways can help the system “hover” stably just short of the breakdown point.
Even if they were smart enough to delay drivers a little longer, long-waits anger drivers – who don’t realize that speeding up the meters means they’ll be delayed even longer. Further, on-ramps may not have enough storage space for longer queuing. That means an angry mayor will complain to the DOT that vehicles are blocking the cross street, and motivate them to run the meters faster to the detriment of the freeway itself. Thus during collapse, perhaps there are 1,000 vphpl who arrive at five, but won’t fit through. Where do they go? Many spill over to parallel arterials, but most just stack back for miles on a mega-billion asset designed to carry them at 65 mph.
Capacity vs. Demand vs. Throughput
Say you have four lanes southbound. Your maximum throughput, or “capacity” is 4 x 2,200 = 8,800 vehicles that could in theory travel southbound in one hour. This Capacity is a number that will not change unless you add more lanes, or until autonomous vehicles are able to increase it by reducing the safe following distance. Now say you have 10,000 vehicles arriving between 5 and 6 pm that want to get on your 8,800 freeway. This is your “Demand,” currently at 2,500 vehicles per hour per lane. But as soon as a few more try to get on than will fit, your “throughput” will drop from 2,200 to 1,500 or below, which means that 4 x 1,500 = 6,000 vehicles will succeed in getting through between 5 and 6 pm, even though up to 8,800 in theory could have, and 10,000 were trying. The 4,000 that didn’t make it have overwhelmed parallel paths, or remain parked on the freeway and will finally get through later.
At this point, engineers will declare a capacity problem, and probably add a fifth lane if they can. 5 x 2,200 = 11,000 as the new fixed value for capacity. Given that there were only 10,000 to serve, 10k / 5-lanes = 2,000 vphpl, which is below the 2,200 mark, so it will work without delay – for a little while.
Latent Demand vs. Induced Demand
In the first month after the fifth lane, drivers who were on parallel paths come back, and the full 10,000 who wanted to use the freeway will be able to, where only 6,000 got through before. Engineers call this “Latent Demand,” – vehicles that were already out there somewhere last month, and have now come back to the freeway because it is working again. But over the next year, “working again” becomes a factor in many daily decisions: “I was shopping close-by, but now I’ll shop further away” – so local shops end up closing as regional “Power-Centers” are all that can survive. Or “I was taking transit, but now I’ll drive again.” But perhaps the strongest, longest-lasting effect is that as people buy and sell homes and change jobs, the fact that the freeway is working again becomes part of their decision in how far away from work they’re willing to live. And as more people adopt far flung lifestyles, because the added capacity enabled them to, the overall amount of driving increases, negating much of the benefit of the added capacity. This additional driving that would not have occurred otherwise is the essence of “Induced Demand.”
So with the combination of latent demand, induced demand, and plain old population growth, what was 10,000 vehicles is soon enough 12,000 arriving at five. The resulting stop-n-go is back, and lanes are again inefficient and able to serve just 1,500 vphpl. So with 5 lanes, that is 7,500 vehicles that make it through between 5-6 pm, even though almost 11,000 could have made it, and 12,000 wanted to. Engineers seem convinced that just one more lane will finally do the trick, but this time it just can’t be done without truly herculean construction.
Humorous Analogy: The Solution! vs. The Solution?
Let’s compare an inefficient freeway to a car that loses power just when you need it most. Say your car runs great until you reach 65 mph, then you suddenly lose power and it sputters along at 30 mph. Your mechanic tells you there’s an electrical problem causing it to fire on just 5 of 8 cylinders. You’re not a mechanic, but you know you need 8-cylinders of power, and there seem to be two ways to get there: 1) Figure out what’s causing the problem so you can run on all 8-cylinders, or 2) live with the electrical problem and instead replace your V8 with two V6’s so that when this happens again you’ll be running on 8-of-12 and not just 5-of-8. Brilliant?… NOT!
Obviously it is far cheaper to just get it going on all 8. But this absurdity is not so different than the scenario that plays out with our freeways. Say you’ve got an 8-lane freeway and it runs great on all 8 until it collapses into stop-n-go, and then it only runs at 60-70% efficiency – basically sputtering along with about 5 lanes of throughput just at the time when we most need the full 8! So we take it to the mechanics – the engineers – for their diagnosis and recommendation, “There are just more cars here than can fit, so your options are 1) find a way to prevent everyone from dumping in all at the same time so your existing 8-lanes will function like 8-lanes, or 2) go ahead and build 12-lanes so that you’ll have the 8-lanes worth of throughput that you need during stop-n-go.”
Showing it this way, it seems obvious we’d want to be more efficient. But up to now the best way to get just enough to go do something else at 5 pm has been a proposal to charge all users a modest fee, figuring that all it takes if for 5% or so to do something else to avoid the fee. Ok, but that means 95% are going to end up paying the fee, and it’s been politically unacceptable for politicians to tell their constituents that they’re basically voting to impose a 5 pm tax on the freeway.
Hence we are tempted by the mega-billion “solution” because it’s “free” as in “you can pay with obscured taxes at the pump.” Or “it’s free for you, because your grandkids will still be paying off the bond long after you’re gone.” Can you imagine double-decking two sputtering V-6’s in your Honda Civic just so you can have one good engine’s worth of power?! Fortunately there are finally some politically viable ways to stay at near-optimal throughput with systems known as Managed Motorways.
Managed Motorways, 101
So what do you do when there are more vehicles entering the system than the system can handle, but you can’t (or shouldn’t) expand the system by creating more lanes? Most areas try to add fixed-guideway transit, improve alternative routes, and convert one or two general purpose lanes to High-Occupancy / Toll lanes (HOT lanes). There is nothing wrong with these alternatives, but they become most attractive only when the politically favored option becomes unattractive. In other words, these alternatives are fundamentally an admission that there’s a big problem not getting solved.
However there is another potential strategy – stop letting too many vehicles enter the freeway! At face value, it sounds un-American – a political third-rail. “I have a right to free, unfettered access to the freeway, and if it’s failing, then spend whatever it takes to make it bigger!” But a new incredibly cost-effective concept known as “Managed Motorways” has proven to be politically possible in Australia, and will likely soon debut in Colorado and Utah.
The logic goes like this…
Say your freeway is already metered and the average wait time is 3-minutes, but the meters do not recognize they’re letting too many vehicles on. So after waiting 3-minutes, a vehicle enters the mainline and is again delayed 30-minutes, inching along for miles at a snail’s pace. Throughput has then dropped from 2,200 vphpl to maybe 1,500 vphpl or lower, which is 70% efficient or less.
Measure mainline flow and if it looks like it’s about to break down, slow the meters. Also interconnect meters to talk to each other, so they can spread the load across many on-ramps.
Design computer algorithms to optimize flow, enabling the system to “hover” at close to breakdown without actually breaking down (perhaps at 95% of the break-down point).
This will cause the average wait at ramp meters to increase, perhaps to 4, 6, 8, or in extreme cases maybe even 10 minutes. The folks will be looking for their pitchforks.
Pacify the brewing mob by creating before/after simulation videos, and “rice through a funnel” demonstrations, to help people appreciate that while the ramp wait may be longer, their overall trip will be very much shorter. (See original 2007 Washington DOT video, and Utah’s recent video)
Since delay at on-ramps will increase, you may not have enough room to store more vehicles. Widen the ramp to 2, 3, 4, or even 5-lanes as necessary to detain vehicles as they await their turn, in order to prevent the ramp from blocking the cross street (Figure 3).
Optional: Install a tolled bypass lane on the ramp, where a driver can elect to bypass the 6-minute wait for a $2 fee. Also allow buses, or other qualifying vehicles to bypass.
Install messages at decision points stating “$2 or 6-minutes”. Short trips will then likely avoid the freeway. (Values are merely examples).
Figure 3 Sometimes on-ramps need to be widened so that Managed Motorway systems have a place to store vehicles. Otherwise they get “stored” in a long parking lot formerly known as “the freeway.”
Recall from Figure 2 in Part 1, the “Easily Possible” system is designed to hover at 95% and by definition is able to support 2,100 vphpl. However recall that demand was 2,500, and hence 400 are still not getting through. First, note that 400 is much better than the 1,000 before. But still, what happens to these 400? Some are diverting to parallel arterials, but not as many as before. Generally, they are being stored at various on-ramps, waiting their turn to enter, where before even more were “being stored” on a multi-mile parking lot.
For locations that already have ramp meters, this could be considered a just minor evolution of the system, because it is just turning independent meters into interconnected meters. It’s a smart thing to do and something the public probably thinks is already being done, given that everything else in the world seems to be interconnected. The Australians have achieved 25% more vehicle throughput during peak periods, a 50% reduction in overall travel delays through most corridors, and 20% to 50% crash-reduction rates along some segments.
Utah’s Citizen-Driven Managed Motorways “Shared Solution”
In Utah, the State has proposed to build a new freeway in northwest Davis County as a relief-valve for congestion on I-15. While it would serve as a relief valve at peak hours, the new freeway is projected to get relatively little use as far as freeways in urban areas generally go, because it accesses comparatively little developable land. In the view of a group of concerned citizens, it would be too much to spend for an under-utilized freeway that would exacerbate sprawl, consume farmlands and wetlands, impact existing development, and make this bedroom county even more of an exporter to Salt Lake County.
So in 2014, these citizens started developing a “Shared Solution” alternative which aimed to achieve similar reductions in delay and VMT in congestion, through many small projects that each share in a larger overall solution. They hired Metro Analytics to help them create project ideas, which included:
Innovative Intersections to reduce arterial congestion;
Protected bikeways to attract people out of their cars
Greater connectivity to relieve I-15 and arterials;
Reduced-fare transit, and upgraded transit, to attract drivers away from congested corridors.
Mixed land uses, and strategies to attract more jobs into the area, to reduce the need to commute to Salt Lake.
And a brand new concept proposed by Metro Analytics which they termed “fail-preventive ramp metering” to ensure that I-15 could perform well without the need for widening nor for a second freeway.
Shortly thereafter, it was discovered that the fail-preventive ramp-metering was not a completely new invention. It has recently been implemented by VicRoads in Australia, and is known as Managed Motorways there. Simultaneous with this West Davis Environmental Impact Statement, UDOT launched a 2015 study to determine the general costs/benefits of Managed Motorways. Where it was already a suggested Shared Solution project in Davis County, and with evidence from Australia that it really is a good idea, UDOT and the local MPO adopted Managed Motorways into their Regional Transportation Plan for all freeways – a very positive outcome of the citizen-driven Shared Solution process!
The region has tentatively adopted a politically-tolerable 6-minute maximum ramp-time. I-15 in Davis would hit that limit and likely go higher unless a 5th lane is added in each direction by 2040. UDOT weighed the pros and cons of adding this 5th lane vs. adding the relief-valve freeway, and seem to be concluding they’re still better off building the new freeway anyway. They do plan to include many suggestions from the Shared Solution process, including making it more like a parkway, and constructing Managed Motorways on I-15.
How Managed Motorways Can Replace “HOT Lanes”
When freeways became unbearably congested and further widening was difficult, many professionals recommended congestion pricing as the ideal strategy to get things moving again. It is true that pricing would clear up congestion and help it “hover” just below the breakdown point at 95% of maximum throughput. But today’s congestion pricing models generally require 100% of drivers to pay at least a nominal fee to drive at five. Sadly that flies in the face of “an apple-pie American institution” – free access to freeways, and thus it has proven politically infeasible so far. Many drivers consider it double-taxation (gas tax + an additional peak period tax), and there is also a concern that freeways will become “Luxury Lanes” – excluding those who just can’t pay extra.
So instead, a political compromise has emerged where one or two lanes are managed through variable tolls and vehicle occupancy requirements, while the majority of lanes remain free, but still painfully slow and highly inefficient in terms of hourly throughput. These “HOT lanes” (high-occupancy/toll lanes) are generally acknowledged by professionals as better than everything being a disaster. HOT lanes are also a baby-step toward a day when perhaps the public will accept a different American institution applied to the entire freeway – the free market, where we “buy our way out of congestion” instead of “building our way out.” Pricing gives you a choice of traveling for free before or after peak periods, or just paying for a scarce commodity – a slot on a peak-period freeway – the same way we pay more for peak-hour electricity, or peak-hour dinner on a Friday at 7:00 pm.
So HOT lanes have been “the best solution for the moment,” but unfortunately they can also be very expensive because it often requires “building a freeway within a freeway” in order to safely separate high-speed traffic from low-speed traffic. Another negative is that HOT lanes, while fast, also often end up with well below ideal throughput on the paid lanes because the goal is often Level of Service C rather than higher throughput of LOS D or E. When it comes to bang for the buck, a letter grade of D or E is much preferred over “getting a better grade.”
So how can Managed Motorways eliminate the need for Managed Lanes? If you better manage access in the first place, then ALL of the lanes on the freeway will run well, and there will be no real incentive for anyone to pay to use a HOT lane. Vehicles will be packed in like sardines, but they will be moving! They may be a little below 65 mph, but they’ll be near-maximum throughput, which is the highest possible return on the public investment in the freeway.
Observations of Inefficiency: Riverside CA’s SR-91 HOT Lanes
I stumbled upon these photos and bar-charts of SR-91, which I believe was the nation’s first serious HOT lanes project. As an engineer indoctrinated in the fact that freeway lanes can carry 2,200 vehicles per hour per lane, the green bar chart caught my eye first. Their before/after data apparently showed that when all lanes were free, their throughput would drop as low as 800 vphpl – a value that stunned me because that is only 36% of 2,200. That is about the same as an at-grade arterial street that is impeded by stop lights. But just as surprising was the “success” of getting the HOT lanes up to 1,600 vphpl.
I thought, “Ok, that’s a big improvement, but why aren’t they getting up to 2,100? Then looking at the photos of prices and lane utilization, it becomes clearer that the HOT lanes are not necessarily attempting to maximize throughput, but may instead be maximizing revenue, or perhaps not focused on maximizing anything. Maybe they’re just aiming for Level of Service C or something, which is a little too much elbow room in my opinion when there is clearly a disaster going on in the free lanes.
It makes sense to maximize revenue in order to pay for the expense of separating fast from slow lanes. But what is not clear to me is whether this “$2.50” is actually the right value to maximize revenue. Maybe $2.00 would not only help democratize both cost and throughput, but also potentially result in more revenue since there would be more users? If the goal is to maximize throughput, then the toll should be set low enough to attract about 2,100 vphpl into the lanes, hovering just below the breakdown point.
How Managed Motorways Can Include Free Market Pricing
An aspect of HOT lanes that many professionals like is the revenue that comes from those who purchase passes, and also the anticipation that a “pay for great service” desire will eventually result in enough people supporting full-fledged congestion pricing because they will come to prefer paying a little to save a lot more. Managed Motorways, as described so far, can successfully keep things flowing and also remain completely free for everyone. But it will induce demand (next section), meaning that it too will eventually stop working.
But while it can stay free and egalitarian – where teenagers on their way to the mall wait along-side a panicked single-mom who is late for a job interview – what if there were options to pay? Suppose you need a 3-lane ramp in order to store vehicles…
Lane 1: “Full cost, but no time lost”
Lane 2: “Half-price, half as nice”
Lane 3: “No pay, but most delay”
Research from the SR-91 HOT lanes, the longest-running system in the country, shows that over 80% of low-income drivers favor the option to pay. They tend to use the free lanes, but they will pay, usually once a week, in cases where the cost of being late exceeds the cost of the fee. An example they note is late fees charged by daycare centers which exceed the cost of the toll.
In a HOT lane system, it may be impossible to have something between “free vs full price” because fast lanes alongside slower lanes would create a safety hazard. However at an on-ramp, each lane can be priced differently so that people can easily choose whichever lane fits the needs of the moment.
Managed Motorways Will Increase Sprawl and Induced Demand
Our “horizontal elevators” (free freeways) make it easy to live far from work, easy to afford large lots with little money, and necessary to have vast swaths dedicated to parking and driving. There are many good things about suburbia such as the ability for so many to own a spacious home where they don’t have to “enjoy” their neighbor’s music, and they have the ability to have land for gardening, children, hobbies, etc. But low-cost land is also offset by high-cost commuting, a household’s need for a car for seemingly every soul over 16, and the high-cost of providing ever wider freeways and ever more basic infrastructure to the ever expanding urban footprint. There is rapid loss of valuable farmland, and it probably takes more fossil fuel to run an expansive city than a more compact one.
Regardless of one’s position on whether low-density has more bad than good or more good than bad, a basic fact is that free access to high speed freeways tends to result in larger urban footprints (lower densities) and more miles driven. While Managed Motorways do not rely on adding lanes as the means of restoring high speeds, the result is still high speeds, and that in turn will result in faster farmland consumption and more VMT. Thus much of the regained capacity will soon be consumed again in part by induced demand. And induced demand eventually creates pressure to either increase the ramp delays to politically difficult levels, or add lanes yet again.
For those desiring more sustainable development, there is a mechanism that will allow high speeds, but also restrain sprawl, increase infill, dampen induced demand, support transit, and encourage shorter trips. Sadly this “magic bullet” has been labeled with two words that people hate: “Congestion” and “Pricing.” So far congestion pricing is 100% infeasible because it sounds like an additional tax designed by greedy politicians that will require 100% of peak-hour users to get out their wallet.
In truth, congestion pricing is really just a free market strategy for matching demand to available supply, to ensure the system will work when it is not practical to increase supply. As useful as that may be in theory, it is still useless as long as it remains politically infeasible. However, we have an idea for a potential “tweak” to traditional congestion pricing that could make it much more politically possible, and that’s the topic of the next section.
Modify “Pricing” for feasibility via “Congestion Credits”
Fast and Free is in the Past: From the 50’s to the 80’s, freeways were both fast and free! – Or at least when they bogged down, it was easy to make the fast again by simply filling shoulders and medians. But from the 90’s to today, a huge share of urban freeways became seriously congested and cannot be easily expanded. Thus conditions have changed and we can no longer have both fast and free at the same time – or at least “free” comes with extremely high construction costs, which is a strange way to define “free”.
Today it’s Free, but It’s Not Fast: As they say, “time is money.” When freeways bog down, the price we pay is in time. If we are unable or unwilling to construct new capacity, and at the same time do not provide an option to purchase premium service with money, then 100% of drivers will pay with time – a quasi-Soviet system that is “free for everyone, but not fast for anyone.” For teenagers rockin’ out on their way to the mall, paying with time is fine. But for anyone desperate to get to an interview, a hospital, or an airport, paying with money is an option they desperately wish they had.
Fast, but not Free, Doesn’t Pass: Congestion pricing as historically envisioned seems to have the opposite problem. It would be “Fast for everyone, but free for no one”. That would keep teenagers off the freeway at 5-pm, (usually a good thing), and it is an appropriate user fee to charge those who use freeways at peak hours for the cost of expansions that were necessary for peak hours. But the drive at five crowd is just too large and flexes their “no-way” political muscle when asked to pay.
Thus HOT lanes so far have been the only thing offering a choice to travel fast when we’re willing to pay, or to tolerate delay when we’re not willing to pay. But as noted earlier, this political “sweet spot” increases construction costs (to safely separate fast lanes from slow lanes), and it also means that in addition to being slow, the majority of lanes sadly have terrible throughput.
Congestion Credits = Fast & Free again, for Most! Metro Analytics proposes a new “Congestion Credits” strategy that blends the positive aspects of Managed Motorways, HOT Lanes, and distance-based Congestion Pricing, to counteract the negatives of each. The result should be a system that will not collapse into stop-n-go (except by unpredictable incidents), can dampen sprawl and induced demand, has the option to pay with money for premium service, but also has an option for most to pay with time if they’d prefer.
This idea requires very little construction to implement on existing freeways (a major advantage over HOT lanes). And where congestion pricing requires 100% of people to pay for a 5 p.m. freeway slot (politically impossible), this system can be free to many and maybe most users at 5 p.m., making it more politically practical.
Here are the basics behind “Congestion Credits.”
Mainline Transponder Towers: Set up transponder towers every few miles spanning all lanes. Traffic flow sensors keep track of mainline flow rates. Price is zero if flows are not in jeopardy (less than about 1,900 vehicles per hour per lane) but as demand climbs higher, increase the price on all lanes as necessary to hover around 2,100 vehicles per hour per lane (95% of max, which is 2,200). All vehicles must pay if they pass under a tower, but they can pay with “credits” if they have enough built up (see below). If a vehicle drives through with no means of payment, just send a bill to the address associated with the license plate.
Fail-Preventive Metering at On-Ramps: Ramps have two to five lanes so that vehicles can be delayed for four to eight minutes without stacking back so far that they would block the cross street. Also create a toll bypass option. Use signs at decision points and via mobile apps that say “$2.00 or six minutes” so that drivers can decide whether to use the freeway or do something else. (Same as Managed Motorways).
Get credit for waiting! All storage lanes on the ramp also read transponders, but lanes with the longest waits will pay you for waiting! Suppose Lane 1 has no wait but costs $2.00. Lane 2 has a 3-minute wait, and choosing it won’t cost you anything, but you also won’t get any credit and hence will start paying by the mile once you enter the freeway. Say Lanes 3 and 4 have the maximum six-minute wait. When you drive under your tower, it gives you a $2.00 credit. In this case, $2 for six minutes is equivalent to $20 for 60 minutes, or $20/hr – well worth the wait for many people.
Use your credits: Now suppose transponder towers are placed every two-miles, and the 5 pm price to drive under a tower is 25 cents. Your $2.00 credit lets you travel free for 16 miles. That should allow a huge share of drivers to have less delay than before, and still be completely free. If your commute is less than 16-miles, credits could roll over allowing you to travel further for free on another day. This creates an incentive to live close to work so you can travel for free and accumulate credits for those times when you must travel further at peak hours.
Traditional pricing kicks in at some point for heaviest users: Say you live 40-miles from work. That’s your right, but you’re creating an outsized burden on society for the miles of freeway you are consuming relative to what you pay for through general fuel taxes. So your first 16-miles are free, and the last 24 might be free also – if demand at that point has dropped below critical, (and hence no pricing is needed). But if demand is still high, then you would start to pay a little more than those who had shorter commutes because you’re a heavier consumer of critical and expensive freeway space than they are.
Special Situations: To increase political acceptance, you could also give out credits for special situations. Express buses have free passes. Maybe registered car-pools could also obtain free passes. You could offer free or reduced rate passes for motorcycles, low emission vehicles, low income residents, qualifying volunteer service, etc. Maybe monthly passes also, where if you’re willing to pay a certain amount up front you can travel as many miles at peak hours as you want without having to pay any more. It requires local debate on what the special situations should be, but they are possible, and that can help with political practicality too.
“Credit lanes” could also make it easier to deal with vehicles that do not have a transponder. Perhaps in addition to a tower that gives your transponder credits, that lane also has a “credit booth” where you can obtain a voucher with your credits that can be read by mainline towers if you are a visitor and don’t have a regular transponder. Also, where the charge lane (Lane 1) would have a camera to send you a bill for $2.00 if you go through without a transponder to keep track of your bill, credit lanes (3-4 in this example) no not need a camera because the system isn’t charging you anything anyway, and that helps deal with “nosy government” privacy concerns, as there’s an option to avoid cameras.
This is not so different than “fast-passes” vs. regular passes at amusement parks, where you decide how you’ll access the ride before you get in, but once in your experience is the same. Also with such a system, it might be possible to manage a freeway for a hundred years without ever expanding it – if the entrance fee as measured in money, time, or both are correctly matched to what the market is willing to pay at that time for premium service.
HOT Bypass Lanes: Good Idea or Bad Idea?
How did we get “HOT Lanes” in the first place? They started out as honor-system High Occupancy Vehicle lanes, and HOV bypass lanes at on-ramps, to encourage people to form car-pools for commuting. The option to pay a toll to enter the lane as a 1-person vehicle was added later for cases where there was clearly excess capacity in these lanes. As Managed Motorways gains traction in the US, there will surely be calls to once again allow anyone traveling with multiple occupants to use any tolled bypass lanes for free. I am in favor of ramp bypass lanes, but I would make them transponder-only lanes. Why not free for HOVs? I didn’t say that. Given that “free for HOV” is politically popular, I would make it possible for government-sponsored vanpools, registered car-pools, and any other locally-determined qualification criteria to obtain a transponder that will allow them to bypass ramp queues for free or at a reduced rate.
Why not skip the paperwork and just use the honor system as always? Partly because too many are not all that honorable, effectively allowing them to butt-in-line at the expense of all others. That in turn will reduce the revenues that are helpful in paying to manage the system. But in my mind the bigger reason might be that it could make it difficult to implement or transition to the Congestion Credits concept. If you enter a ramp with choices to pay for no delay, or to just wait it out and get transponder credits to help you out down the road, if too many people cheat the system, it could lead to excessively long delays in the credit lanes as optimization algorithms designed to prevent failure can never let the honest ones in if there are too many butting in line. Requiring transponders is cleaner – and can still be compatible with rewarding those who are trying to do what they can for sustainability and to save money.
ROI – This is Worth a Try! It’s Worth $Billions of Tries!
Managed Motorways should make it possible to raise throughput from below 70% to as high as 95%, and raise peak speeds impressively. Adding in pricing and credits can dampen any induced demand. So how much is all of that worth? And how much would it cost to achieve it?
To get a ball-park figure, we used Salt Lake’s travel demand model and EDRG’s TREDIS transportation economic benefits calculator to estimate that this 2.5 million region would gain about $50-billion in societal benefit over 25-years. “Societal Benefit” includes things like an impressively higher Gross Regional Product because shippers, companies, and high-value individuals are able to recover a lot of time, much of which they turn around and use for higher productivity. It also includes the value people place on time recovered for such things as getting home for dinner on time, being able to see your kid’s performances, and other quality of life issues. It further accounts for likely reductions in accidents, and some environmental benefits such as lower emissions.
The modeling compared conditions under their 2040 Regional Transportation Plan, which tries but fails to “build its way out of congestion,” to a similar scenario where all freeways in the region operate at no worse than Level of Service E – obtainable through any of the above systems that are able achieve 95% efficiency.
How much could it cost? For this Salt Lake experiment, construction would primarily consist of electronics to better measure system flows and to interconnect ramps, as well as widening perhaps as many as 50 on-ramps to create more storage capacity. No one has yet tried to pencil out what that could cost to implement across the whole region, but a generous figure for construction and ongoing maintenance might be $1-Billion. That would be a 50-1 Return on Investment! Try to find ANYTHING in the transportation realm with a regionally significant effect that can get much above a 10-1 ROI.
Spend less, get more! Further, these benefits can be obtained at a cost savings relative to other options. For example, suppose the region is planning to spend $10-billion on freeway expansions that their own models say will still be sputtering at 70% or less efficiency due to stop-n-go. But if they spend $1-billion on this, they can likely achieve better overall regional statistics without having to spend anywhere near $10-billion on general freeway expansion.
Addressing Common Concerns
People have a lot of worries about Managed Motorways, especially if there is talk of congestion pricing. Common concerns include:
“Forcing people off the freeway will lead to car-mageddon on parallel streets!”
“Automated vehicles are right around the corner, so we won’t have to worry about this.”
“This is a double-tax, and we already pay enough as it is!”
Here are several thoughts on addressing common concerns:
It could make parallel arterials better off! When a freeway fails, a LOT of drivers converge on parallel arterials hoping for a better way through. It is true that pricing, or even extra delay at on-ramps, will also motivate some drivers to use parallel arterials, but both freeways and parallel arterials could still end up better off. How? It’s counterintuitive, but the math suggests this may be the case. If a system in failure has 60% throughput, then restoring it to 95% means you will be moving 35% more. Where did this “more” come from? In part from those already avoiding the freeway but who were always willing to pay with either ramp delay time or money, had that been an option. Thus managing the system to 95% throughput means that even if a few are avoiding the freeway, even more are coming back because they can now fit through.
Automated Vehicles: There are likely several decades to go before automated vehicles will be common enough to travel bumper-to-bumper together and hence raise the maximum freeway throughput from 2,200 vphpl to something higher. And what about all of the non-automated vehicles here now and yet to be purchased? They’ll be on freeways too for a long, long time. Do we want to miss out on the benefits of congestion-free systems while we wait? Further, automated vehicles also have potential to exacerbate sprawl and create induced demand, unless they are implemented with a VMT tax or some kind of induced demand dampener. If they finally come on-line, they will increase the threshold of breakdown, but there will still be a limit to maximum throughput. When that limit arrives, Managed Motorways will still be helpful.
Managed Motorways can be free: Managed Motorways does not actually need to charge any money in order to work well, because you can “pay” with time delays at on-ramps rather than the actual money of HOT lanes or congestion pricing. And paying with time at on-ramps costs less than what you pay in time parked on a failed freeway. However, if the on-ramp is the only “payment point” of either time or money, the result will induce demand. That is because it favors long trips, since you do not have to pay with any additional money or time delay once in the system, and that means over time people will choose homes even further from their jobs. Induced demand means that eventually you may need to increase ramp wait times to politically challenging levels, or revert back to building your way out again.
Tolled bypass lane is a voluntary purchase: A tolled bypass lane on the ramp is an optional element to allow some users to pay with money if they’d rather not pay with time. But it still favors long trips since once in the system there is no need to pay again. Though government may be collecting the money, it cannot be fairly claimed as a tax or fee, because it is voluntary. This is simply selling something of value – time savings.
Congestion Pricing is a system management tool, and not a typical tax: The purpose of pricing is to prevent system collapse. There is no reason to charge any money or delay anyone at on-ramps unless the system is threatened with imminent failure. From Economics 101, if demand exceeds supply, and you’re unable or unwilling to increase supply, then you must raise the cost in some way so that some people – and it doesn’t take many – will voluntarily choose to do something else. Congestion Pricing requires payment with money, while Managed Motorways and Congestion Credits allow drivers to pay with time. Since distance-based pricing affects all users, I guess you can call it a tax if you must. But unless you can think of another way to dissuade enough drivers, then time or money are your tools.
Revenues collected are simply the means of achieving that. But for political good-will and sustainability, it is wise to channel revenues toward projects that provide alternatives, such as making parallel rail or express bus routes free at five – a strategy that might more than double ridership for a very marginal increase in cost. Pricing is a more perpetually sustainable system than merely managing on-ramps, because it dampens induced demand and creates a mechanism whereby demand can always be matched to supply. It is less voluntary than a HOT lane or a tolled bypass lane because 100% of users would have to pay at certain times. It might still be considered “semi-voluntary” for many and probably most trips, as people still often have a choice to travel before or after the fee, or travel by transit or alternative routes, but many will still find that on any given trip they really had no practical choice but to pay the fee. That’s where Congestion Credits kicks in!
Congestion Credits increase political feasibility: We believe our proposed congestion credits system improves upon traditional congestion pricing because it creates a mechanism for many to pay the mainline distance-based transponder fees with credits for the time they spent waiting at on-ramps rather than paying with money. That makes it free for anyone who can complete their trip within the credits they have. But there will still be some constituency, hopefully small, who make regular, very long 5-pm trips and would end up paying at least some money. I am not sure how to get around that.
This “Double-Tax” could reduce net taxes for everyone: Say today’s gas tax is 30-cents, and there’s a serious proposal to increase it another 20 cents to pay for system expansion – some of which will be for Big Dig construction, and none of which is predicted to fully eliminate freeway stop-n-go. Pricing need not represent a net increase in taxes. What if the region instead reduced today’s gas tax to just 25 cents, made up the difference through pricing, and obtained huge gains in “new freeway capacity” just by choosing to run at 95% efficiency rather than 60-70%? 25 cents + pricing would be a tax cut for many and maybe most voters – because in truth most voters are probably not regularly on a freeway that would be priced at 5 pm.
Also, because pricing allows the existing freeways to operate well without “Big Dig” expansions, the region avoids raising the gas tax to cover forgone projects. Thus while long-distance 5-pm drivers are indeed paying above the regional average, they and everyone else are also paying less overall with this “double-tax” relative to the expensive construction scenario, where gas taxes would have been raised substantially on everyone.
But even if we could create a great evidence trail to prove this as a fact, there would still be great concern among the public who hasn’t seen or doesn’t trust your analysis. The majority of voters know they’ll end up on freeways at 5 at least sometimes, and the power-users will not like being targeted to pay above the average, even if they understand that they’re more heavily responsible for the cost of freeways.
There is also something psychologically tangible about being reminded every day that you’re paying 25-cents every two miles, vs. a gas tax that is baked into the price and hence doesn’t seem like you’re paying anything at all. It will take a lot of powerful, concise messaging to convince enough people that even if they’re paying above average, they’re still paying less than they otherwise would be and getting better service at the same time.
Political Feasibility via Performance Based Planning
As we approach an era where there are more needs than funds, there is a desire to elevate projects with the highest Return on Investment, or highest Benefit / Cost Analysis (BCA). This BCA approach has MPOs, DOTs, etc. all scrambling to define what “Benefit” means, and how to weigh benefits in one arena against benefits in another.
Many of the typical performance arenas are shown below. Using baseball as the analogy, Managed Motorways and Pricing that adopts our Credits system are very likely to “hit home runs” relative to traditional ways to spend similar amounts of money. There are some “strike outs,” but they all can probably at least get base-hits in Political Feasibility, where striking out would mean it never gets tested.
|Performance Area||HOT Lanes||Managed Motorways||Pricing w/Credits|
|Base Hit – Better than nothing!||Triple – Entire system performs well at less cost.||Home Run – less induced demand and ability to raise price if nec = more ability to remain perpetually stable.|
|Capital & Maintenance Costs||Base Hit – may offset some expansion||Triple – New capacity without expansion||Home Run – Can limit expansion needs for decades.|
|Safety||Not sure – fast next to slow issues?||Double – stable flow improves safety.||Home Run – stable flow and reduced VMT both improve
|Economic Vitality||Base Hit – minor benefits for some. Trucks still stuck||Triple – All users save time, and time = money.||Home Run – Same as MM, but sustainable for longer.|
|Land Use (sprawl/farms)||Home Run – High delay = less sprawl||Strike Out – Same effect as widening freeway||Triple – Promotes short trips, transit, and Activity
Center infill. No credits = Home Run, but also infeasible.
|Multi-modal||Triple – Fixed guideway attractive if fwy not moving||Base Hit – Fixed guideway less attractive, but Exp.
Buses move well.
|Home Run – Use revenues to make transit free and get
huge ridership gains. Also people use transit to avoid additional cost of driving.
|Net Taxes||Strike Out
– Fast/slow separation is expensive; pressure to widen remains
|Triple – Reduces need to widen, which saves on taxes||Home Run – Longer effect = less construction need =
lower taxes even with the “double tax” (two sources of
|Political Feasibility: At least a base hit is required!||Home Run – Proven to have high favorability!||Home Run? – Minor variation of ramp metering, and appears like it will work.||Base Hit? – Where pricing has been a strike-out,
credits could help pricing get on base. But it will still be very challenging.
In my estimation, HOT lanes offer high return on investment relative to standard, unmanaged freeways; a Managed Motorway has even higher ROI than HOT lanes; and Pricing w/Congestion Credits would have the longest-lasting benefits across more categories than even Managed Motorways. So the real question then is how to bring these new high-performance systems to a point where people are just enough willing to give it a try and see how they like it.
Managed Motorways is pretty simply an upgrade of standard ramp metering, so the biggest public hesitation will be a need to appreciate how longer ramp delays equates to shorter overall time and better safety. This education seems relatively easy to achieve. Pricing w/Congestion Credits will be politically challenging not only because it involves a new payment, but it’s a bit hard to understand and will take time for enough people to be convinced it’s worth taking a chance. But fortunately it can be added to a Managed Motorway later. So even if land consumption “strikes out” with MM, land may win out in the end if MM helps make Pricing down the road more feasible. Tolled ramp bypasses and congestion credits counteract claims of double-taxation by creating more voluntary ways of upgrading your experience, and ensuring free options remain for many if not most drivers.
Where Congestion Pricing has always been politically infeasible, the new emphasis on performance frameworks may also provide some political cover: “Don’t blame me, blame it on the things you said you care about! All your performance metrics suggest that Congestion Pricing has the highest return on investment for the things you care about, and the new Credits concept helps ensure that many can still travel for free or at a greatly reduced price even for long trips!”
Summary and Conclusion
The first region that can get its freeways running reliably at 95% efficiency without collapsing into stop-n-go could see a return on investment of 50 to 1, and possibly much more than that when accounting for foregone “Big Dig” projects that no longer need to be constructed. The benefits are worth billions at a regional level, and likely trillions if adopted nationally, in part due to higher economic productivity that comes from reliable high-speed freeway mobility. And though high speeds normally induce demand and lead to sprawl, that can be ameliorated with distance-based congestion pricing. And pricing, in turn, would create an incentive to live close to work, take transit, and a host of more sustainable actions.
But great as pricing may be in theory, sadly it has proven politically infeasible thus far. Our idea for “congestion credits” could allow many trips to still being free, which should increase the political feasibility of congestion pricing. This “Advanced Formula” for freeway optimization could be a win-win hybrid of Managed Motorways, Congestion Pricing, and Congestion Credits that allows payment with time or money. If some rendition of this proves to have just enough political sugar to help the medicine go down in at least one region, it could create a tipping-point of innovation where what has been impossible suddenly becomes inevitable world-wide.
While the effort to convince the public of this may still take decades and may only inch forward in baby-steps, perhaps now there are more baby-steps we can make, and have a better overall story to tell.
Michael Brown is president of Metro Analytics based in Salt Lake City, Utah.
Related Links and References
SR-91 Project Info Sheet: http://www.sr91project.info/media/upload/xsk7jlovqu.pdf
Managed Motorways article, Salt Lake Tribune, July 2015: http://www.sltrib.com/home/2738864-155/ramp-meters-could-be-coming-to
Robert W. Poole, Jr., Gaining Public Support for Freeway Congestion Pricing, Reason Foundation, Policy Study 401, April 2012.
Doug MacDonald – Rice and Traffic Congestion, YouTube, April 20, 2007: https://youtu.be/8G7ViTTuwno
Chen, Chao and Varaiya, Pravin. The Freeway-Congestion Paradox. Access. Number 20, Spring 2002
Transportation Economic Development Impact System (TREDIS) http://tredis.com/