There is no reason we should be adding highway capacity in proportion to highway growth. Trying to build our way out of congestion is a dead paradigm [1]. Times have changed [2]. We know that GDP grows at a faster rate than population, and that GDP and VMT (vehicle miles traveled) are closely correlated.
Keeping up
If we want to keep vehicle delay (roadway congestion) under control, we need to add lane miles at the same rate we add VMT. Seems plausible, given that GDP grows at the same rate as VMT. But only if we keep up--once congestion on a road reaches a certain level, people choose different routes, different times, and different modes. And then when we widen the road again, the people just change back, the 'triple convergence'. So once a road is congested, it will never be uncongested again.
Even if we double the size of the road?
Nope. People will just divert from other times, other routes, and other modes, to the new wider and faster route. And in the long-term, people will react to higher driving speeds by moving further away from things. When we first built the interstates, there was no congestion, and everyone moved out the central cities--until about 1970, when the 'Urban Transportation Problem' of roadway congestion happened, and people stopped moving, because there was too much congestion. And so we began to widen, and have never been able to stop.
Digging a Hole
When you did a hole, the deeper the hole goes, the harder it gets, because each increment of depth you've already created is harder than the last--you did the easiest stuff first, when you started digging. Adding lane miles is like that. At first, we just paved the existing street. Then we eliminated on-street parking. Then we took the parkstrip. Then the sidewalk. Then the front-yards in front of houses. But at some point, to widen the road, you have to buy the whole house. And for the biggest roads, that's where we are at. TexDOT is proposing to take an entire subdivision (?!!) to widen it's central freeway. UDOT is going to take a row of houses along Foothill Boulevard, at $350k each.
Rising obligations
The second half of the problem is maintenance. Roads don't stay built--we have to maintain them. Every lane mile requires a certain amount of money to stay in good condition. The amount of money varies by the age of the road--costs are low at first, but rising, with a big lump sum every 45 years or so, when the road has to be 'rebuilt'. The more roads we have, the more we have to maintain. [3].
So when we try to build our way out of congestion, we're taking on not only the cost of new roads, but the cost of maintaining all roads we've already built. GDP/capita is increasing as fast as VMT/capita, but not as fast as VMT/capita+maintenance per capita. [4]
The reality is actually worse, because the money from gas tax doesn't exactly track GDP. Gas tax revenue is a function of VMT and fuel efficiency. Fuel efficiency rose for decades, so the amount of gas tax collected per VMT fell over time, and we refused to raise it. (Why the Highway Trust Fund is broke, btw). [5]
In conclusion, we can't build our way out of congestion, because the cost of keeping up with congestion grows faster than our ability to pay for it.
#----------------------------------------------------------------------------------------------------------------
[1] If you are non-metro area with slow population growth, you can do it! But you'll got broke paying for it, 30 years hence. Check out Chuck Marohn of Strong Towns for more on that.
[2] Gas used to cost five cents a gallon. I suspect it was not the 'Freeway Revolts' that put an end to US roadbuilding, but rather the Arab Oil crisis (1973, 1978(?)). Check the spikes in 1973, and again in 1978. Oil prices jumped 20x in 10 years! Destroyed demand for driving as well as the major input into road construction.
[3] It's worse if we try to 'stretch' the road when the repairs come due. Water gets into the cracks of the roads, the freeze-thaw cycle levers the asphalt apart or washes out the dirt underneath the road.Then average annual cost of maintenance per lane mile increases.
[4] How do I know? It's a bit complicated, but I used Excel. I found some numbers on construction costs per lane mile, and cost of the annual maintenance costs of roads, on a per lane-mile basis (and co I had one column of GDP growing at a certain rate, one column of VMT growing at the same rate, one column of lane miles growing at the same rate. And a cost per lane mile, growing at the same rate as GDP and VMT. Then I added a cost for maintenance, based on the numbers I found, times the cumulative lane miles built. Then I summed the cost of new lane miles and the cost of lane mile maintenance. Then I graphed them, to see if GDP kept pace with the total cost of roads. It didn't. Over time, the cost of roads outstripped the increase in GDP. When this happened varied by what I used as my inputs, but the dynamics are inevitable: the combined growth in the cost of lane miles and the cost of lane-mile maintenance outgrew GDP.
Mathematically, it would be something like X + .00025X > X
[5] Ah, the ever-reliable "Only because the Mass Transit Account!" argument. I'll address that elsewhere. Today, I'll simply note that: I would love a per ticket charge on every transit ticket to fund mass transit capital spending on buses and guide-way. The Highway Trust Fund has been a reliable slush fund to promote road building for the past 100 years. I'd love transit to have the same thing.
I'd also like mandatory mass transit parking with every new development, and exclusive 'transit-only' lanes, the way cars have. Fairs fair, right? Oh, and complete elimination from liability from manslaughter--that'd be cool. And heck, I'd take as many dedicated lane miles as proportional to the number of person-miles traveled. If mass transit is 3% of the total, 3% of the lane miles seems fir.
Keeping up
If we want to keep vehicle delay (roadway congestion) under control, we need to add lane miles at the same rate we add VMT. Seems plausible, given that GDP grows at the same rate as VMT. But only if we keep up--once congestion on a road reaches a certain level, people choose different routes, different times, and different modes. And then when we widen the road again, the people just change back, the 'triple convergence'. So once a road is congested, it will never be uncongested again.
Even if we double the size of the road?
Nope. People will just divert from other times, other routes, and other modes, to the new wider and faster route. And in the long-term, people will react to higher driving speeds by moving further away from things. When we first built the interstates, there was no congestion, and everyone moved out the central cities--until about 1970, when the 'Urban Transportation Problem' of roadway congestion happened, and people stopped moving, because there was too much congestion. And so we began to widen, and have never been able to stop.
Digging a Hole
When you did a hole, the deeper the hole goes, the harder it gets, because each increment of depth you've already created is harder than the last--you did the easiest stuff first, when you started digging. Adding lane miles is like that. At first, we just paved the existing street. Then we eliminated on-street parking. Then we took the parkstrip. Then the sidewalk. Then the front-yards in front of houses. But at some point, to widen the road, you have to buy the whole house. And for the biggest roads, that's where we are at. TexDOT is proposing to take an entire subdivision (?!!) to widen it's central freeway. UDOT is going to take a row of houses along Foothill Boulevard, at $350k each.
Rising obligations
The second half of the problem is maintenance. Roads don't stay built--we have to maintain them. Every lane mile requires a certain amount of money to stay in good condition. The amount of money varies by the age of the road--costs are low at first, but rising, with a big lump sum every 45 years or so, when the road has to be 'rebuilt'. The more roads we have, the more we have to maintain. [3].
So when we try to build our way out of congestion, we're taking on not only the cost of new roads, but the cost of maintaining all roads we've already built. GDP/capita is increasing as fast as VMT/capita, but not as fast as VMT/capita+maintenance per capita. [4]
The reality is actually worse, because the money from gas tax doesn't exactly track GDP. Gas tax revenue is a function of VMT and fuel efficiency. Fuel efficiency rose for decades, so the amount of gas tax collected per VMT fell over time, and we refused to raise it. (Why the Highway Trust Fund is broke, btw). [5]
In conclusion, we can't build our way out of congestion, because the cost of keeping up with congestion grows faster than our ability to pay for it.
#----------------------------------------------------------------------------------------------------------------
[1] If you are non-metro area with slow population growth, you can do it! But you'll got broke paying for it, 30 years hence. Check out Chuck Marohn of Strong Towns for more on that.
[2] Gas used to cost five cents a gallon. I suspect it was not the 'Freeway Revolts' that put an end to US roadbuilding, but rather the Arab Oil crisis (1973, 1978(?)). Check the spikes in 1973, and again in 1978. Oil prices jumped 20x in 10 years! Destroyed demand for driving as well as the major input into road construction.
[4] How do I know? It's a bit complicated, but I used Excel. I found some numbers on construction costs per lane mile, and cost of the annual maintenance costs of roads, on a per lane-mile basis (and co I had one column of GDP growing at a certain rate, one column of VMT growing at the same rate, one column of lane miles growing at the same rate. And a cost per lane mile, growing at the same rate as GDP and VMT. Then I added a cost for maintenance, based on the numbers I found, times the cumulative lane miles built. Then I summed the cost of new lane miles and the cost of lane mile maintenance. Then I graphed them, to see if GDP kept pace with the total cost of roads. It didn't. Over time, the cost of roads outstripped the increase in GDP. When this happened varied by what I used as my inputs, but the dynamics are inevitable: the combined growth in the cost of lane miles and the cost of lane-mile maintenance outgrew GDP.
Mathematically, it would be something like X + .00025X > X
[5] Ah, the ever-reliable "Only because the Mass Transit Account!" argument. I'll address that elsewhere. Today, I'll simply note that: I would love a per ticket charge on every transit ticket to fund mass transit capital spending on buses and guide-way. The Highway Trust Fund has been a reliable slush fund to promote road building for the past 100 years. I'd love transit to have the same thing.
I'd also like mandatory mass transit parking with every new development, and exclusive 'transit-only' lanes, the way cars have. Fairs fair, right? Oh, and complete elimination from liability from manslaughter--that'd be cool. And heck, I'd take as many dedicated lane miles as proportional to the number of person-miles traveled. If mass transit is 3% of the total, 3% of the lane miles seems fir.
No comments:
Post a Comment
And your thoughts on the matter?