Everyone Knows We Have a Traffic Problem

I'm in Sarasota, Florida, this week, where an election was just held to choose two new city commissioners. Eight candidates ran, and in nearly every interview and campaign mailer I saw from any of them, one word was at least as prominent as any other: "traffic." The lead vote-getter in particular ran as a slow-growth candidate advocating for tighter scrutiny of new development in the city, linking this stance to concerns about traffic congestion.

New downtown development: Traffic scourge or scapegoat?

This dynamic should be familiar. If you've waded into local politics around growth and development almost anywhere, "traffic" is a drum that citizen activists love to beat. It makes sense: traffic is one of the most visible aspects of quality of life that local government is in a position to affect for better or worse. It matters.

So let's talk about traffic.  It's a truism that people in every city believe they have a traffic problem (just like everyone believes they don't have enough parking). But for all the talk of traffic problems down here, I've heard comparatively few viable solutions, and I suspect part of that is because we so rarely bother to really define the problem.

The most frequently offered policy changes I see called for here are a blanket call to "stop overdevelopment," or to charge developers concurrency fees—establishing a pot of money that can somehow be used to mitigate traffic problems resulting from population growth. How? Good question.

Other, more wrongheaded solutions, pop up now and then too. A local newspaper's editorial board recently suggested turning a major downtown street into a "raised flyover" in a disturbing echo of elevated freeway projects that have ruined American neighborhoods for decades, and in some places still threaten to.

Here's the Fort Myers example the local editorial referred to. Look at that lovely parking crater in the adjacent blocks of downtown. No thanks:

A more sophisticated discussion of what a "traffic problem" actually means, what causes it, and what will or won't solve it is sorely needed, here and in hundreds of similar cities. Parts of this discussion are specific to Sarasota, but many of the issues and solutions are very generalizable. Inasmuch as "everyone knows we have a traffic problem" here, Sarasota is Everytown, USA.

What a Traffic Problem Isn't

The default assumption for many is that we have a traffic volume problem—"There are too many cars on the road." Or it's that we have a traffic delay problem—"All the other cars on the road are slowing me down." This assumption is pervasive in our public discourse about transportation, fed by propaganda from the engineering lobby, for example the TTI Urban Mobility Report and the INRIX congestion index. It's dangerous, because it presumes a specific set of solutions. Namely, if the problem simply boils down to too many cars and not enough capacity, either we widen roads and make more room for more cars (at colossal and ruinous expense), or we somehow put a stop to development and make all those people go live somewhere else.

Freeway traffic in Atlanta, GA. Source: Wikimedia Commons.

Neither is the solution, and the above premises are wrong. In fact, the evidence is that there is a very inconsistent relationship between development density and travel time; very long commute times affect those who live in the spread-out suburbs of Houston and Atlanta as well as in transit-rich New York and Chicago. And as for delay—lost travel time due to congestion—this is a slippery notion, because the baseline for measuring "delay" is arbitrary, and its actual cost in people's lives is hard or impossible to measure, since people don't actually value their travel time the way these simplistic models presume they do.

So no, I don't believe we have a problem with too much traffic. I do believe Sarasota and many other cities actually have a problem with both the nature and the distribution of traffic. But this prescribes a different set of solutions, because the problem ultimately comes down to the qualitative experience of driving here, not a quantitative measure of travel time or delay. Perhaps we need a "Traffic Frustration Index" to replace the Traffic Congestion Index.

To lay out why I believe "traffic frustration" is the real issue, let's step back and examine what different things we might actually mean when we say that traffic is a problem.

What a Traffic Problem Is: Three Possibilities

Let's get this out of the way first: "too many cars on the road" doesn't mean anything. How many is too many? 1,000? 10,000? 100,000? Over what area? Over what span of time? Are the cars sitting still? Are they free flowing? Are they arranged in a Stonehenge-like roadside tourist trap? A traffic problem can't just mean that there is traffic; it means that people's lives are being negatively affected by traffic. So I propose three ways in which that could be the case.

  1. It takes too long to get places you need to or want to go.

  2. The time it takes to get places you need to or want to go is too variable or unpredictable.

  3. The experience of getting places you need to or want to go is unpleasant because of traffic conditions.

It's important to determine which of these things are actually true in order to get at solutions that are affordable, effective, and without harmful unintended consequences.

Explanation 1: It takes too long to get places you need or want to go.

This feels like an arbitrary concept from the get-go. What does "too long" mean? How long is long?

It turns out that the history of cities offers a pretty compelling answer to this question: more than an hour round trip. This is called Marchetti's constant, and it was offered up by the Italian physicist Cesare Marchetti in 1994 as a bound on city size: a city would only grow to the extent that residents using the prevailing transportation technology of the era could commute to and from their work in about half an hour each way. People were not willing to spend more than that amount of time traveling. Marchetti's constant is remarkably accurate in predicting the size of ancient Greek and medieval European villages as well as modern American metropolises.

Ancient Athens: city size determined by 30-minute walking distance.

Kansas City streetcar system, early 20th century. City size determined by 30-minute transit distance.

Atlanta metropolitan area, current-day satellite photo from Google Maps. City size determined by 30-minute driving distance.

This interactive map of travel distance by ZIP code in the U.S. reveals a remarkably consistent pattern if you pan from major city to major city: average commute times are rarely much under 20 minutes or over 30 minutes.  (The exceptions are mostly fringe exurban or rural areas within commuting distance of big cities—especially where those big cities have an inflated cost of housing, as in coastal California for example.)

20 minutes to 30 minutes. This is true of large cities, and it's true of smaller cities. It's true of compact cities, and it's true of sprawling cities. The Census Bureau's 2015 American Community Survey reports that the mean commute time for the U.S. is 27.2 minutes: pretty close to Marchetti's 30.

Let me emphasize this once more, because it's an important point that most people get wrong: denser development does not mean longer travel times. It doesn't because in a compact city, while you might move more slowly between destinations, your destinations are also closer together.

How do local conditions in Sarasota compare? Favorably, it turns out:

(Data source: American Community Survey, 2011 through 2015 five-year estimates)

Not only is the average Sarasota resident's commute significantly shorter than that of the average resident of the U.S., Florida, or Sarasota's own suburbs, average commute time here has also barely budged over the past few years. It has grown by only 1.5%, while national commute times grew 7.4% in the same period.

Now, there are some potential problems with this. It's a yearly average, in a place with a high seasonal population, so these numbers don't reflect seasonal conditions. It only reflects commuting and not other trips. However, on the whole, it still doesn't seem that drivers in Sarasota have it particularly bad when it comes to travel time.

Of course, ‘It takes too long’ could simply mean, ‘It takes longer than I’d like it to take.’

Of course, "It takes too long" could simply mean, "It takes longer than I'd like it to take." I, too, plan to ask Santa for a personal teleporter this year. Now, anyone is welcome to lobby their government in self-interest for a freeway that just happens to begin a few blocks from one's house and end a few blocks from one's workplace. Write to your City Commissioner. Go for it.

But the realm of serious policy solutions requires that we not take seriously claims of hardship based on the idea that one is entitled to a certain travel speed or time between one's desired destinations, no matter the adverse effect of a wider or faster road on the neighborhoods it passes through, or on the solvency of the municipal budget.

Plus, are we going to build our way to faster travel by expanding roads? If we try, we're going to encounter this pesky little thing called induced demand, itself an implication of Marchetti's constant, and a phenomenon well established in the transportation planning literature and understood by transportation agencies. If people, on average, are willing to tolerate trips up to about a certain length, then adding capacity won't reduce trip lengths on a citywide scale, for two reasons:

One is that in the medium to long term, land use will compensate. People will be willing to live farther from work, or from non-work destinations that are important to them, knowing they have a high-speed connection to those destinations. The incentive for outward sprawl, and more and more traffic as a result, will continue until travel times restabilize at about the same level as before.

In the short term, travel behavior will also compensate. The economist Anthony Downs is one of many scholars to describe induced demand. In his 1992 book, Stuck in Traffic, Downs coined the idea of the Principle of Triple Convergence. Here's how it works:

The Principle of Triple Convergence is best explained by a hypothetical example. Visualize a major commuting freeway so heavily congested each morning that traffic crawls for at least thirty minutes. If that freeway were magically doubled in capacity overnight, the next day traffic would flow rapidly because the same number of drivers would have twice as much road space.

But very soon word would get around that this road was uncongested. Drivers who had formerly traveled before or after the peak hour to avoid congestion would shift back into that peak period. Drivers who had been using alternative routes would shift onto this now convenient freeway. Some commuters who had been using transit would start driving on this road during peak periods. Within a short time, this triple convergence upon the expanded road during peak hours would make the road as congested as before its expansion. Experience shows that peak-hour congestion cannot be eliminated for long on a congested road by expanding that road’s capacity if it’s part of a larger transportation network.

The idea of simply paving our way to reduced travel times should thus be a non-starter. If average travel times were well above the Marchetti constant, we might have a structural problem where they're too long and want to address that through either transportation capacity (allowing faster travel to destinations) or development density (bringing destinations closer to each other). This doesn't seem to be the case here or in most of America, however.

Traffic jam, Maputo, Mozambique. Source: Wikimedia Commons.

Explanation 2: Travel time is too variable or unpredictable.

 It's more likely that most people's traffic woes are not really about absolute travel time, but the predictability and consistency of travel time. To discuss the implications of this possibility, let's differentiate between the two types of variability: predictable and unpredictable.

Predictable variability is, for example, knowing that there's a daily rush hour at which traffic speeds will be slower and travel times longer. If you know this and it's consistent, you may be able to plan your routine around it. Leave earlier, run errands later, try a different route. If you don't have flexible work hours, you may not always be able to avoid rush hour, but you can hope that other drivers can and will.

Another source of predictable variability that Florida cities are especially bedeviled by is seasonal. Every winter, an influx of retiree "snowbirds" swells Sarasota County's population by 25%, and the effect on traffic is as you would expect. Mid-March here is the single busiest time of year on the roads because not only are seasonal residents still around, but spring break sends many Northerners flocking south.

FDOT makes much of its traffic data publicly available, including seasonal adjustment factors to convert "off-season" traffic counts to "in-season" ones. The latest figures for Sarasota suggest that average daily traffic from March 6th to March 14th—the busiest week of the year—is 1.31 times that of the period from August 30th to September 12th, when the traffic is most sparse.

This introduces a legitimate planning dilemma: should we build our roads to smoothly accommodate the anticipated traffic during the busiest weeks of the year? Much like sizing retail store parking lots with Black Friday in mind rather than the other 364 days of the year, this makes no fiscal sense. It makes no environmental sense. Any private company that operated this way would go bankrupt; this is why airlines raise their fares at Christmas, instead of owning enough planes to meet that peak demand without having to do so. It's reasonable that year-round Florida residents should have to suck it up and expect a certain amount of congestion during peak season. But how much inconvenience should they be expected to endure?

(Image: Light traffic on the bridge to the beach in May, outside of peak tourist-and-retiree season. Source: Google Street View.)

To limit the pain from this predictable, annual spike in local travel, the most effective toolkit is on the demand side, not the supply side. We can provide alternatives to cars: public transit, water taxis, free shuttles for tourists. We could implement congestion pricing (whoa! heresy!) on bridges to the barrier islands to disincentivize driving to the beach at busy hours. Cycling and walking—if and where safe infrastructure is provided for it—can also pick up some of the slack from driving: after all, we're fortunate in Florida that in the peak-traffic winter months, it's typically dry, sunny, and in the 70s!

Unpredictable variation, though, is a more vexing traffic problem than predictable variation.

"I hate taking the cats to the vet, because I don't know if that drive is going to take 20 minutes or 45 minutes." That was one of the answers I got when I asked a few friends to examine my list of interpretations of what a "traffic problem" means. When the time a drive will take is that unpredictable from one day or hour to the next, the reason is usually that there's a possibility of encountering gridlock.

This problem of traffic jams, however, has less to do with capacity or with the overall amount of development or population growth. It has more to do with choke points: how that capacity is distributed throughout the street network.

Suburban planning in the U.S. creates its own congestion problems by its heavy reliance on a hierarchical network: one which funnels traffic onto relatively few arterial routes instead of dispersing it more evenly throughout a complete grid. This is analogous to a watershed in which runoff flows into creeks which flow into ever-larger rivers—and rainfall upstream can cause catastrophic flooding downstream.

A complete grid, in which there are many equally-attractive routes from A to B for drivers to choose from, actually does a great job of dispersing traffic and preventing congestion. Wide, high-speed arterial roads and expressways, on the other hand, are congestion magnets. The bigger the capacity, the greater the slowdown in conditions of high demand. This curious fact owes itself to a phenomenon once again described by Anthony Downs in Stuck in Traffic:

Nearly every vehicle driver normally searches for the quickest route, one that is shorter or less encumbered by obstacles (such as traffic signals or cross-streets) than most other routes. These direct routes are usually limited-access roads (freeways, expressways, or beltways) that are faster than local streets if they are not congested. Since most drivers know this, they converge on such ‘best’ routes from many points of origin.

The problem is that during the peak travel hours on weekdays, so many drivers converge on these ‘best’ routes that they become overloaded, particularly in metropolitan areas. Traffic on them eventually slows to the point where they have no advantage over the alternative routes. That is, a rough equilibrium is reached, which means that many drivers can get to their destinations just as fast on other roads. At times, the direct road may become even slower than alternative streets, and some drivers eager to save time will switch to them. Soon rough equality of travel times on both types of route is restored at the margin. The opposite happens if travel becomes slower on alternative streets than on the expressway.

This is the crucial flaw in a road network dominated by wide multi-lane arterials (which are often stroads): those arterials enjoy a speed and travel time advantage at off-peak hours, but under rush-hour conditions they can become more and more congested until they offer no advantage at all over slower, less direct alternate routes.

I did a little experiment to test my intuition that this is exactly what happens in Sarasota at weekday rush hour. At about 4:00 p.m. on a Tuesday, I drove from east to west across downtown on the principal arterial, Fruitville Road. The distance was exactly 1 mile, the speed limit 30 miles per hour. There are 8 stoplights in this stretch. I timed the trip: 5 minutes, 46 seconds. A lot of that time was spent waiting at lights, 3 of which I caught on red.

Route A: Fruitville Road

Then I looped back and drove west to east across downtown on Main Street. The distance was the same, but Main Street is two-lane with angled parking, ample pedestrian activity, and a 15 mile per hour speed limit. This stretch featured a stop sign, two roundabouts, and three traffic lights. The travel time? 5:51—almost identical to what it took me on a road with double the speed limit.

Route B: Main Street

I repeated the experiment with an analogous north-south pair: one four-lane arterial (Washington Boulevard) with a 35 mph speed limit and 6 stoplights; followed by one parallel two-lane urban street (Orange Avenue) with a 25 mph speed limit, 3 stoplights and a roundabout. In this case, the narrower, "slower" urban street dramatically overperformed the high-capacity arterial, allowing me to go a mile in 4:37 versus 6:39.

Route C: Washington Boulevard (US Highway 301)

Route D: Orange Avenue

There was no congestion on the lower-speed streets, and they were much more pleasant environments—I spent less time stopped, more time moving, and had more street life and scenery to look at.

A map of the routes I tested (Image from Google Earth with graphics added)

 An Aside: How Long is Too Long?

As an aside, because I suspect someone is thinking it at this point: Is 5 minutes an unreasonable amount of time to drive a mile at rush hour? 12 miles per hour seems pretty slow. But this is downtown: travel speeds across most of the city are much faster.

At 4 p.m., it takes 35 minutes to drive from Lakewood Ranch, a rapidly-growing suburb, to Lido Beach on the Gulf of Mexico, a trip that requires passing through downtown Sarasota. If the time to get through downtown doubled, it would take 40 minutes. Would the difference between 35 and 40 minutes deter you from making that trip? Would it severely affect your quality of life?

On the flip side, we could widen roads or remove accesses through downtown, destroying untold millions of dollars in land value and economic activity in the process, and reduce Lakewood Ranch beachgoers' trip time from 35 minutes all the way down to... 31? 32? If this seems worth it to you, I really don't know what to say.

Pedestrian activity, downtown Sarasota.

Finally, we get to Traffic Problem Manifestation #3:

3. The experience of getting places you need or want to go is unpleasant because of traffic conditions.

 The notion that traffic is unbearable in a place is often a firmly held gut feeling by residents who haven't dug into any data or even timed their own trips. I suspect that a lot of this gut feeling stems from unpleasant driving experiences which are highly salient—basically, we remember vividly the times when getting somewhere was a pain in the butt, while the unremarkable, painless journeys are quicker to fade from our memories.

Thus, what if we could significantly improve drivers' happiness in a city not by making their journeys faster, but by making them more pleasant? Reducing the aggregate Traffic Frustration Index?

Here's a far-from-exhaustive list of factors that might lead to unhappy driving experiences, and what we can do about them:

1. Stop-and-go traffic

It's no fun to sit still. Most people I've asked tell me they'd rather drive at a steady 25 miles per hour than experience alternating periods of going 40 and going 10. This is a huge part of why routes B and D in my experiment above were way more pleasant than routes A and C. What did B and D have in common? They were on two-lane urban streets, on which traffic flows at a steady but slow rate (and on which you'd be a maniac to try to go 40 miles per hour) and roundabouts preclude the need to come to a complete stop at every intersection. Compared to the full-on stroad experience, this kind of driving is a breath of fresh air.

Urban roundabouts keep things moving slow but steady.

2. Fear of a collision

Main Street was lively with pedestrians when I drove down it. I didn't worry about hitting them at all, let alone hitting another car. Why? Because I wasn't going very fast, and could react to anyone who stepped out in front of me. The worst that can happen at 15 miles per hour simply isn't that bad. The worst that can happen at 40 miles per hour is horrific.

See this guy crossing the street mid-block? Slow speeds mean neither he nor I was afraid I would hit him.

3. Design speed / travel speed mismatch

Look at the two photos below. They're both theoretically of congestion—in that the free flow of cars is impeded by other cars. But if you're like me, being in the line of cars on the left looks rage-inducing, while the photo on the right doesn't look all that stressful.

Why is that? One explanation is that roads and streets have a design speed: a speed that is comfortable to travel in free-flowing conditions, based on the lane width, presence or absence of visual obstructions, presence or absence of on-street parking, and other features. Driving at or near the design speed just "feels right" to the average driver. Driving well below the design speed feels wrong. It makes you antsy. It feels slow.

Going 15 miles per hour on the street on the right doesn't feel as subjectively "slow" as going 25 miles per hour on the street on the left, because of the visual cues that tell you what the design speed is, i.e. what speed you "should" be going.

4. Boredom

Which line of cars above would you rather be sitting in? A or B? I'd pick the one with visual stimulus, a lively environment, and some greenery any day over the one that's just... a line of cars.

Would you be more relaxed queued up at this stoplight?

... or this one on a parallel route a couple blocks away?

Would you rather have traffic move through your neighborhood on one of these?

... or on three or four of these?

Recap: What kind of traffic problem do we have? 

So we've got three different interpretations of a "traffic problem," and different sets of policy and design solutions for each:

The Problem

The Cause

How to Address It

1. It always takes me too long to get places.

Travel is too slow for my liking, or destinations I want to visit are too far apart for my liking.

Transportation policy makers probably can't systemically change this, because of induced demand. Individuals can, by living closer to work / recreation / shopping.

2. I get stuck in occasional severe congestion.

Predictable peaks

Demand management: provide alternatives, use congestion pricing.

Unpredictable gridlock

Distribute traffic more evenly across the grid to prevent breakdowns in flow.

3. The experience of driving is unpleasant.

Stop-and-go traffic, dangerous conditions, monotonous environments, mismatches between design speed and travel speed.

Traffic calming—roundabouts, narrower lanes, landscaped medians, on-street parking, etc.—all designed to create a slower but more regular and safer flow.

Problem 1 is mostly unsolvable by transportation planners and probably not really a problem if you live in a place where commute times are under about 30 minutes, as most people do. Problems 2 and 3 boil down to the qualitative experience of driving more than to the actual quantitative delay experienced.

So maybe the problem with traffic isn't that there's not enough capacity, or too many cars vying for the same streets. Maybe the problem with traffic is that we've designed both our individual streets and our street network in such a way as to make driving a far more unpleasant experience than it needs to be. And the places in which we're stuck sitting in traffic are places where nobody in their right mind would simply enjoy being in the first place.

The good news is, if our traffic problem is one of driver frustration and not one of travel time or total volume, we can improve things much more easily than we can alter absolute travel time or total volume. Changing the latter two things is a Sisyphean task, given the reality of induced demand and Marchetti's constant. But fixing the frustration piece? The toolkit is right in front of us, as evidenced by the experience of driving on well-designed two-lane urban streets that already exist. In this context, we drive at a moderate but steady speed, an experience which doesn't feel chaotic or unsafe, and which offers lively streetscapes and scenery to look at.

This realization leaves us free to advocate for high-quality, compact development—development which creates destinations that are pleasant to spend time at on foot, and that reduces the need to hop in a car in the first place—without adopting the misguided fear that such development is going to unleash Carmageddon.

High-quality, compact development, brought to you by slow streets.

Everyone knows (Your City's Name Here) has a traffic problem. Now let's have a more sophisticated and forthright conversation about what does and doesn't make sense to do about it.

(Top image from Wikimedia)


Related stories