Ok, this is weird. Washington’s transportation department thinks that vehicle travel is going to go up up up. See, for example, this graph…
Total Vehicle Miles Traveled 1980 – 2030 (projected)
(Miles in billions)
The solid red line to the left represents historic traffic volumes—technically, vehicle miles travelled, or “VMT” in planner-speak. The dashed line to the right is a prediction of future VMT growth.
But look: there’s something awfully suspicious about these predictions.
If you close one eye and tilt your head to one side, you’ll see that the dotted line is much steeper than the solid line. In the real world, traffic growth has slowed way down in recent years, and looks like it’s on the verge of flattening out. A closer look at the numbers (pdf link) tells the same story: the recent slowdown in vehicle travel has been pretty remarkable. As oil prices have soared, the growth in vehicle travel has slowed; and travel per capita has actually fallen.
But the reality of rising fuel costs hasn’t sunk in to the planning department. Instead, they’re predicting a massive, immediate increase in the growth rate of traffic, starting next year. Yep, if you believe their predictions, traffic will grow four times faster over the next 5 years than it did over the last 5 years, at a pace 50 percent quicker than the long term average.
Huh?
It gets worse. Measured per driver, VMTs have been trending downward since the early 1990s. Today, it’s about the same place it was 2 decades ago. But the transportation department is predicting an immediate reversal of the recent declines, with rapid upscaling of our transportation habits—starting, once again, next year.
Vehicle Miles Traveled per Licensed Driver
1980 – 2030 (Projected)
I could go on—the WashDOT is claiming anomalous growth in VMT per capita and per vehicle as well. Of course, all of these predictions could turn out to be correct. That’s the thing about the future—you can’t know what’s going to happen until it actually happens. But based on recent trends, both in traffic growth and oil prices, my money’s against WashDOT.
Whoops!! Actually, that’s not true. My money—that is, my tax dollars—is actually being spent, right now, based on the assumption that WashDOT’s predictions are accurate. You see, these projections aren’t just an academic exercise. They’re being used to argue for new roads, increased traffic capacity, and all the trappings of a car-dominated transportation system. And by building new roads to accommodate anticipated capacity, we could actually be helping to make those predictions come true. New roads, for example, could facilitate sprawling development on the urban fringe, which in turn will make car travel even more of a necessity than it is now.
So I’m left with 2 possibilities. Either transportation planners haven’t caught up with the new realities of rising gas prices, growth management policies, and demographic trends that are reducing our need to drive. Or misguided transportation investments, justified by inflated traffic projections, will create the trends that planners are anticipating.
Either way, it looks to me as if it’s time for some new projections.
Eric de Place
Adding that these projections are basically against the law.That is to say, they’re especially odd in light of the fact that Washington has a new law requiring per capita VMT reductions of 18% by 2020 and 30% by 2035. Given current population projections, the new per capita targets mean that total VMT in 2030 would be almost the same as what it was in 2005. But I’m stealing my own thunder from a future post…
pwk
I know these numbers aren’t used in designing WSDOT projects in Central Puget Sound. WSDOT uses the Puget Sound Regional Council model, which includes PSRC’s assumptions about land use and VMT. If PSRC’s VMT numbers are inflated, maybe we don’t need that nasty congestion pricing after all.Has anyone bothered to ask WSDOT how they use these VMT projections?
Clark Williams-Derry
Fair question, PWK. I haven’t asked.
Stacey W-H
This model was used in Olympia when locals tried to convert a section of Old 99 (Capitol Way, near downtown Olympia) from 4 lanes to 2, adding a center turn lane, bike lanes and pedestrian amenities. Although current usage data for the road suggested it was feasible, the traffic engineers rejected the plan because the projections showed the road would eventually exceed capacity. I argued at that time that the projections shouldn’t be used but it’s a hard case to make when faced with 50 years of historical data. I argued exactly the point of this article: that the projections will become self-fulfilling if we don’t build the infrastructure to change it. Cheap fuel is a thing of the past, and the model should be as well.
mahalie
When I first glanced at your article I thought the chart might be for global miles traveled, not Washington. In which case I was going to remind you that a lot of developing countries, China in particular, are showing explosive growth in vehicle ownership as their economies emerge. But then I saw this was for Washington. Between the lightrail opening in 09, Seattle’s enviro freaks and gas prices…these predictions just seem silly.
J.A.K.E.
Its all about rates (percents) and real numbers, dude. The rate of growth of VMT is slowing (95% in 27 years vs. prediction of 54% in 23 years) but the total VMT continues to grow: more people (born, reaching driving age, living longer, immigrating) all increase total VMT even if we all start driving the same or even less. The rate of growth decreases, but not the growth. Check the projected rate: it is one-half the rate for the last 27 years. But all those new drivers continue to drive up the total VMT. And the new laws? May they be as effective as those making speeding, littering and murder illegal. Good luck.Now if you could create a solid economic depression, or a plague, you might be able to reduce total VMT.
Bryn Davidson
At the Dynamic Cities Project (a Vancouver BC based NGO) we’ve been working on alternative transport demand scenarios that incorporate potential impacts from oil depletion and/or climate legislation. In these scenarios we model future changes in vehicle efficiency, alternative fuels, and reductions in demand from high fuel prices – as driven be emerging oil scarcity or future carbon taxes. The result? In most cases it appears that we have as much road capacity today as we are likely to ever require, or be able to use (i.e. peak roads). In the case of freight trucking, it could be that we’re actually going to be facing substantial declines in VMT over the next 10-40 years.This is not to say that an increase in total VMT isn’t possible, but it is looking increasingly unlikely that such an expansion will manifest to the extent that we should spend billions of dollars on road expansion that could be better invested elsewhere.From our point of view, linear demand extrapolations are very misleading (or at least increasingly out of touch) and any long term infrastructure investments should be planned and tested using a scenario based approach.
Chris Bradshaw
Driving slower has its own rewards, like reduced noise caused by the higher impact of air particles striking the surface, which is related to the energy required to get through them.The reference to ‘pace cars’ is to David Engwicht’s concept.Find this information at:http://www.lesstraffic.com/Programs/Pace%20Car/Pace%20Car.htm
Barry
I agree with the emphasis of J.A.K.E.’s comment: “The rate of growth decreases, but not the growth.”The only measure of success in terms of climate change, peak oil, pollution or any of the rest is total usage.The total consumption of all fossil fuels is rising. Tripling the gas price has not cut total consumption of gas. Quadrupling the jet fuel price has not cut total consumption of jet fuel. A sharp run up in coal pricing has not cut coal usage. In fact the opposite is happening…both locally and globally.Correctly modeling the future demand as Bryn’s comment talks about has proven to be very hard to do. I’d like to think we are at “peak roads” but i see no sign of it based on past behaviour or current trends.I think the problem comes back to our inability, so far, to predict the “elasticity” of fossil fuel demand. “Elasticity” is not a linear factor even in the average. It depends on the price sensitivity across the full spectrum of consumers of a resource. It also depends on how much of that resource each wealth-level consumes. Finally it depends on cost of alternatives.Here are some trivial examples to highlight what i think is happening: 1) LINEAR DEMAND. Ten people each use 10% of a resource. They have the following dollars to spend per unit of product X: $1, $2, $3, $4, $5, $6, $7, $8, $9, $10. As price of X rises there is linear 10% drop in demand for each dollar rise. Easy to model based in initial trends.2) DISCONTINUOUS ELASTICITY. Ten people each use 10%. Their available dollars to spend per X: $1, $2, $3 and then $1,000 for the final seven. As price of X rises you get same initial linear demand. But suddenly the pattern stops and elasticity seems infinite. If you try to model the future based on early trends it doesn’t work.3) HYPER-ELASTICITY. Elasticity becomes huge if the people with the most money to spend are also the ones consuming most of the resource. Ten people again. The amount they each use is: 2%, 2%, 2%, 2%, 2%, 5%, 5%, 10%, 20%, 50%. The dollars to spend per X match this: $2, $2, $2, $2, $2, $5, $5, $10, $20, $50. As you can see the price of X needs to go over $10 before there is any significant cut in demand. At that price you have only cut demand 30% but you have also priced 80% of people out of the market. In this “elasticity” scenario, raising the price of X creates widespread hardship long before it leads to significant cuts in X. Could this be what is happening? Are we headed for “carbon riots” and “cap and awe” rather than carbon cuts?We know that wealth is increasingly stratified in our society and around the world. There is data showing strong ties between wealth and fossil fuel consumption. For example, Professor Steven Pacala at Princeton says less than 8% of humanity causes 50% of global emissions. Three billion people compete in the fossil fuel market with $2 or less per day.If we have this “hyper-elastic” scenario our current set of pricing policies is going to fail. We are going to see things like a $3 increase per gallon of gas with no drop in total demand. We will see things like 5% growth in jet fuel demand for every 5% increase in jet fuel price.If so, we will also need tiered pricing or personal caps or some other tool to address the hyper-elastic demand.I’d love to hear from any modelers what assumptions they are using on distribution of consumption by wealth-level. Getting it right seems to me to be the crux of any successful model/policy.
Bryn Davidson
Barry thanks for the comments. You bring up an important point about price elasticity and social justice i.e. will the rich buy up energy to maintain their entitlements at the expense of the poor’s core services (food, shelter etc.). I believe this is already happening. At $100+/bbl oil is priced out of the reach of poorer countries – impacting inflation, disposable income, and the availability of food, among other things. What this phenomenon is doing is hiding the effects of peak oil from the wealthy world as we draw down the reservoir of available ‘demand destruction’ in the third world. After that we’ll draw away demand from our domestic poor, and then from the middle class in economically faltering communities. In terms of modeling, we’ve argued that the availability of oil-based products in US and Canadian cities will track more or less with global production. This is based on an admittedly rough assumption that declining ‘net-exports'(export declines from OPEC and other producers who subsidize rapidly growing internal demand) will be balanced by the extra oil we siphon off from poorer market competitors – at least until the well of demand destruction runs dry and depletion hits us squarely. Given all the uncertainties, we run two depletion scenarios:1. High Case = 1% growth in available oil to 2015, then 1% decline beyond2. Low Case = Flat to 2010, 2% Decline to 2015, then 4% decline beyond (the ‘experienced’ depletion rate accelerates after the reservoir of third world demand destruction is exhausted by, say, 2015)On top of these we bundle sets of assumptions for growth of non conventional fossil fuels (ie coal to liquids), bio-fuels, electrification, efficiency, and population.In the end, you have developed a ‘depletion savvy’ high and low case for total VMT and per capita VMT that you can put side by side with the BAU demand projections, and make the argument:”given these three plausible futures, what’s the most resilient investment in infrastructure I can make?” I think most people on Sightline know what the best answers might be, but my hope in doing this modeling is to provide a bit of substance with which to balance the charts and data of the traffic engineers and highway industry marketers.Bryn