ANALYZE THE CONTROVERSIES

Digging through the archives: let's consider induced demand.  Let's open with Strong Towns's Daniel Herriges, "The Mobility Trap: Why We'll Never Fix Congestion by Speeding Up Traffic."  That post went online before I heard of Wuhan or coronavirus, and at the time it was Donald Trump sucking up to the highway lobby.

After a four-year reprieve, the Texas Transportation Institute (TTI) has once again published its infamous Urban Mobility Report. So if you're seeing a spate of headlines about how the traffic is bad and getting worse, coupled with outrageous dollar amounts that Americans supposedly lose to congestion—like $166 billion in 2017—the source of those claims is almost certainly the TTI report.

This report is pure propaganda for the road-building industry, designed to elicit credulous takes from major news outlets that cement in the public's mind the idea that We Need to Spend More On Infrastructure™. It inflates the cost of congestion in a dozen different ways, by making unreasonable assumptions that don't reflect the real world. It has been debunked repeatedly by a range of transportation policy experts, and yet it just. Keeps. Coming. Back. And without addressing any of the substantive criticisms.

The post provides links to those substantive criticisms.  Go, read, understand.  Some of the substance follows.

Glaring problems with the report include that it assigns the value of time lost sitting in traffic based on prevailing wages, which is a nonsensical practice to economists: the alternative to losing 6 minutes on your commute is virtually never that you could have otherwise earned 6 additional minutes of pay at your job. And very rarely would anyone be willing to spend 6 minutes worth of their wages (say, on an express toll lane) to avoid that congestion—turns out this has been studiedand people will pay, on average, more like $3 an hour, or well under minimum wage.

Another error: the study defines “delay” (congestion incurring a cost) as any time vehicles are moving less than free-flow speed. But on some roads, the free-flow speed is actually greater than the legal speed limit. Drivers simply obeying the law are counted by the TTI as evidence of congestion.

The glaring errors in the report go uncorrected because correcting them would defeat the report’s purpose, which is to make congestion look bad and always worsening. Because others have done such a thorough job deconstructing the report's definitions and assumptions, though, I'm going to attack it from a more root philosophical angle that helps explain why this report is asking and answering the wrong questions.

Yes, and Burma Shave had the right idea about what commuters would do with six additional minutes, years ago, plus it's difficult to attribute "delay" on a road moving at slower than the speed limit without considering how much slower, if at all, travel would be on other roads available for that trip.

Those road improvements might have had some effect on bid-rent curves and land use, but the commuter of today still faces a similar travel time to the commuter of 1982.

Travel time has barely budged: almost everywhere in the U.S., the average commute is between 20 and 30 minutes. This is true whether you're in a huge city or a smaller one, a spread-out one or a dense one, a congested one or a relatively uncongested one. It's true because most people's travel time budget is 20 to 30 minutes for a one-way commute, and that budget is going to inform decisions about where they live, work, and go about the rest of their lives.

Travel distance, on the other hand? On average, we drove almost twice as far each day in 2017 as we did as recently as 1982.

So what do we achieve by building new highways and speeding up travel? We don't actually shorten people's trips; we just enable them to live and work farther away from each other. The results are reflected in the VMT statistics: we drive far more miles than we did a single generation ago. We burn more fossil fuels. We develop more farmland. But it's worse than that: when we prize the ability to go places quickly over the actual value to us of the places we're going, we end up destroying our cities.

I don't know, there might be causes other than highway construction destroying our cities, but that's for a different day.  About the same time that Strong Towns article dropped, Reason Foundation's Robert Poole offered a, shall we say, contrary point of view.  First, it's difficult to do an induced demand analysis without a careful understanding of all the margins along which a motorist might adapt to a new road. "Analytically, they treat parallel arterials and all other roadways in the metro area as a large blob. That leaves them unable to analyze (as opposed to only speculating about) the extent to which drivers shift trips from parallel arterials to the newly expanded urban Interstate (which would not represent new driving, but simply a re-allocation of existing driving)."  That's the first principle of project analysis: any new road will generate traffic (people who previously weren't traveling at all) as well as divert traffic from existing roads.

[Alan] Pisarski suggests looking at why more motorists show up on the expanded freeway: some switch from parallel routes because they find the expanded freeway an improvement over their former route. Those who shift from other travel times do this because it increases their utility to be able to drive at that hour. And the same goes for those who shift from other modes, such as transit. Hence, the expanded freeway “improves and expands choice for both previous and new users.”

But it's risky to bring up the benefit principle, although Mr Poole does:

To be sure, there are locations where the cost of adding freeway capacity may significantly exceed the benefits to highway customers. When that is the case, I agree with Duranton and Turner that the solution is to implement congestion pricing. And in some of those cases, the additional revenue from pricing may be high enough to make the capacity addition feasible.

The extension of "some of these cases" to "treat the roads (and water mains) like productive assets" is straightforward.

A recent Strong Towns column, also by Mr Herriges, expands on the challenges of project analysis.

There are actually two different effects that get lumped together under the heading "induced demand."

Effect 1: In the near term, expanding a road causes drivers to alter their travel patterns to use it more.

Effect 2: In the medium to long term, expanding a major road will alter the pattern of development in a city or region, which itself induces more driving.

The first applies to almost any good or service. The second, though, is fairly unique to roads, and we'll talk below about why.

Pedant's note: the near term effect is what we understand as the law of demand, with people consuming more of something they perceive as cheaper.  Mr Herriges has that: "Economics tells us that, all else equal, when you reduce the cost of something, people will consume more of it. As far as this goes, 'induced demand' is really just 'demand,' and it's nothing unique to roads. Put another way, that's a short run substitution, and we analyze that as diverted traffic.

The medium-to-long term effect is why demand elasticities are higher in the long run.  What manifests itself as an altered pattern of development in a region is the same thing that manifests itself as an altered pattern of reading with the introduction of books on tape or virtual books, and we can come up with all kinds of substitutions as a response to wireless internet or any of the other information technologies.  What the roads do is affect, you guessed it, the rent gradient.  It's probably a vain hope, though, to preserve the existing urban configurations by not building new roads.  "It would be far cheaper to build and live in cities where the people and things we care about are simply nearby to begin with."  Unfortunately, that puts the obnoxious stuff, whether it's a noisy neighbor or a drop forge nearby as well.

But the efforts of the highway lobby to deal with the congestion their unpriced projects produce, as inevitably as Soviet housing policy produces khrushchobas, become increasingly silly.  Consider, for instance, the "continuous flow intersection," perhaps better understood as a "displaced left turn" intersection.  It's the latest attempt to improve traffic flow, which apparently it does by accumulating the left turn traffic on the principal highway (this thing probably can't be installed where two principal highways cross) in a section isolated by two traffic signals, rather than backing up into the principal highway the way the conventional setup does.  Its promoters suggest the extra signals (what do we call them, the home signal and the starter?) actually cause less frustration than the most common configuration.  "Although this design requires that traffic signals be added several hundred feet from the intersection, their added location makes it easy to process a high volume of vehicles. Thus, most motorists stop no more than once going through the intersection." (Good luck negotiating that thing on foot or on a bike, scooter, or skateboard, but I digress.)

Judge for yourself.

Because of the extensive reconstruction involved in providing the slip lanes and other bling, these, like the single-point intersection or the diverging diamond, will likely involve construction delays that future motorists will never get back.