It's Time to Stop Repeating This Damaging Myth About Traffic
If you live in a place where growth is a political issue at all, you've almost certainly heard this refrain. Whether in the endless sunny suburbia of Texas, Florida or Arizona, or the prosperous and growing environs of Denver, D.C. or Seattle, it's easy to encounter someone equipped with the following pieces of supposed gospel truth:
1. “The traffic is terrible.”
2. “The reason the traffic is terrible is because our infrastructure is inadequate to support new development.”
The first of those statements might well be true, in a way. (Emphasis on "in a way"; everyone thinks they have a traffic problem.)
The second of those statements, though, is almost certainly not true. And, even though it's often used by well-intended, thoughtful people, it's a claim that fosters some dangerously simplistic and even harmful assumptions about how growth and infrastructure should work.
An op-ed in my local paper provides a representative example of the "statement #2" type of rhetoric as used by growth-control advocates. The writer identifies a real problem, and one that is at the core of Strong Towns's work: my community, like thousands of others, faces an enormous gap between our identified infrastructure spending needs and the actual resources we anticipate having to meet them. We face this predicament even though we're more affluent than many of our peer regions.
At Strong Towns, we explain this predicament in terms of an unproductive pattern of development that produces more liabilities than revenues. But this writer offers a different, very common diagnosis that is ultimately inaccurate and counterproductive:
"Florida has hamstrung its counties... [and given] developers a big pass: They no longer have to pay for deficient road conditions.... [We face] unbridled development with rapid increases in density and depreciated money for road expansion."
It's an old cliché that when you have a hammer, you treat every problem as a nail. If you define the cause of traffic congestion as "development without adequate road expansion," you have drastically limited your understanding of what's really going on, and you've given yourself only two choices to address the challenges that come with growth:
Build and/or widen more roads, or
Strictly limit development, regardless of what nasty side effects (like inflated housing prices and long commutes) might result.
All because of a misdiagnosis of what is actually responsible for the infrastructure woes common to growing cities.
What Is "Adequate"?
I responded (in a Facebook exchange) to that op-ed writer and others with the observation that residents of the Sarasota-Bradenton metro area, where both she and I live, are at the 95th percentile in per capita driving, according to (somewhat outdated, but the best comparative breakdown I could find) data from the Brookings Institution.
“Nowhere in the United States are we using our roads to capacity. Not even close.”
Here, we average about 13,000 annual vehicle miles traveled (VMT) per capita. The average resident of the greater New York area (including Long Island, Northern New Jersey and western Connecticut) drives less than half as much. Even within Florida there is great variation: metro Miami's per capita VMT is only about 70% of ours.
In this, the problem is certainly not “overdevelopment” or underbuilding of roads. We are overdependent on driving—and driving long distances. If we looked at that 13,000 mile figure and asked, "What would it take to reduce that number by half in the next 20 years," we'd be having a very different conversation than the one most advocates on either side of the growth wars here want to have.
How much infrastructure is "adequate" depends on how we use it. How much of the Earth's surface is required to transport us varies dramatically with how we build our places and get around. Just witness the gap in impervious surface per capita between compact, walkable Europe and spread-out, car-dependent North America.
Our demand—or perceived "need"—for road space is a moving target. By adopting longer commutes at higher speeds, we can double or quadruple our consumption of publicly-maintained asphalt with no real time savings or benefit to quality of life. It's an endless mobility trap, and it's fed by the conventional wisdom that we need to keep expanding our roads to alleviate traffic.
Let me be clear: Nowhere in the United States are we using our roads to capacity. Not even close. Even at afternoon rush hour, most streets are close to empty. Yes, there is traffic congestion in many places. But what actually causes congestion is a combination of isolated system breakdowns, and bottlenecks that have more to do with poor road connectivity and car-dependent development than with a system-wide lack of capacity.
We need to deal with congestion by reducing the amount that we are forced to drive, not by expanding roads.
We've Been Here Before
A number of U.S. states have actually coded into law the idea that infrastructure must be adequate before development is permitted. Florida was a trailblazer in this in the 1980s, with statewide "concurrency" requirements in place from 1985 until 2011. The principle of concurrency means that public infrastructure must be expanded or improved before or alongside—"concurrently” with—any private-sector development (like new homes or businesses) which increases the demand on that infrastructure. Maryland and other places have similar laws and call them Adequate Public Facilities (APF) ordinances.
Typically, the way they work is that when new development is proposed, the developer must analyze the impact on roads and water systems—and sometimes on other facilities such as parks or schools. If the development would push any of those services above a level of congestion that the local government has deemed acceptable, the developer must pay for public "improvements" (usually road widening) to reduce that congestion back into the acceptable range. Otherwise, the development will not be allowed to proceed.
What constitutes “congestion” is up to the local or state authorities, but in the case of roads, local governments typically define it as a minimum acceptable Level of Service (LOS), a flawed metric which enshrines the idea that cars ought to be able to travel as smoothly as possible without delay, even on city streets.
Most slow-growth or managed-growth advocates where I live regard concurrency as a good policy that should be reinstated. They're wrong. In practice, concurrency was an object lesson in unintended consequences from the beginning. The Mercatus Center explains why (boldface emphasis mine):
New development could not take place unless the road network was sufficient to handle the additional traffic the development would bring. Concurrency was very popular, both with antigrowth Floridians who thought that a lack of infrastructure would slow growth and with developers. Developers believed that with concurrency in place, local governments would be pressured into funding infrastructure improvements that would support continued growth.
One of the hopes of the Growth Management Act’s supporters back in 1985 was that concurrency would require local governments to fund infrastructure, but when the financing did not appear, many jurisdictions had roads that were too congested to legally allow more development, so development was increasingly pushed outside the central core of those jurisdictions to the periphery where roads were less congested, thus creating the very type of sprawl that the [state] said it wanted to prevent.
Greater Greater Washington has documented similar perverse side effects of these rules in Maryland. Not only do Adequate Public Facilities (concurrency) laws render it easier to build at the suburban fringe where congestion is not yet as severe as in urban areas. The resulting traffic also has downstream effects on the urban core, further exacerbating congestion there and incentivizing building outward in a vicious cycle: "[m]otorists tend to be like water [and] fill up whatever space is available.”
The Victoria Transport Policy Institute, and many academic scholars (here's one example) also describe concurrency requirements as having contributed to spread-out suburban development by imposing disproportionate costs or barriers on development in already built-up areas.
When applied in outlying areas, concurrency rules can require the construction of more road capacity than is actually needed, or of suburban-type road capacities (i.e. 4 or 6 lane arterials) in rural areas that development hasn't reached yet. This results in shorter commutes for people who live way out in the boonies, which is an inducement to people to keep moving there and developers to keep building there.
People who say, "No new development unless there's adequate infrastructure to serve it" are generally well-intended, but my fear is that line of reasoning ends up just reinforcing an unspoken assumption wherein "adequate" starts to mean "a convenient, uncongested commute from wherever I chose to live." At whatever expense is necessary.
So What Do We Do Instead?
We ask a different set of questions when faced with the uncomfortable effects of growth, for starters.
We recognize that the amount of infrastructure we "need" isn't fixed, or a simple linear function of population growth. Demand for public infrastructure tends to expand alongside its availability. In the absence of prices for the use of infrastructure that reflect its full societal and long-term costs, congestion is an important feedback mechanism that causes us to rein in our consumption.
Uncomfortable feedback is how our places grow resilient, as we make individual adjustments which add up to a collective evolution in how we live and what it costs us. Traffic congestion, for example, leads to avoidance strategies: live closer to work, drive at off-peak times, take transit or walk. Congestion of some public facilities like parks works similarly. I live in a beach town, but our most famous beach is outrageously busy every weekend. And thus many locals go to different beaches, or at different times.
Trying to spend our way out of congestion—this unpleasant but important feedback—is futile. A few years ago, we doubled the capacity of the free parking lot at Siesta Key Beach because it was frequently full... and guess what? It's still frequently full.
This doesn’t mean our cities and counties should never build infrastructure. But we should do it judiciously, incrementally, and in response to obvious present needs, not way in advance of projected future needs—because it will just turn those needs into self-fulfilling prophecy. Fiscal discipline says, “We should make frugal and resourceful use of every bit of public infrastructure we build”—since all of it, even if the developer pays up front, will ultimately have to be maintained at public expense.
The logic of concurrency and "inadequate infrastructure" complaints, on the other hand, says, "If we're experiencing uncomfortable feedback from our use of public or shared resources, government's role is to just keep providing more of the resource—or restrict who can access it—until the discomfort goes away."
In a finite world, that can't be the answer.