How Much Does Congestion Matter?

Michael Lewyn's picture
When Transportation Secretary Ray LaHood's suggested that bicyclists' needs should be accommodated in federally-funded road projects, the road lobby responded with something approaching hysteria. Bill Graves of the American Trucking Association wrote that a more pro-bicyclist policy "would cause an economic catastrophe" by "hinder[ing] the movement of our nation's goods."(1) The road lobby's logic seems to be (a) supporting bicycling reduces funding for roads, which (b) will lead to an increase to road congestion, thus (c) causing an economic catastrophe. 

In this blog post, I'd like to focus on element (c) of that chain of logic- the link between road congestion and economic growth.  Both auto and transit lobbies occasionally suggest (as did Graves) that without more transportation funding, economic life as we know it will end.

If this were true, the most congested regions would be the most economically stagnant ones, and the least congested regions would be booming. But this is hardly the case.  According to the Texas Transportation Institute, the two large regions with the lowest per-capita travel delay are Buffalo and Cleveland (2)- hardly economic powerhouses.

To be a little more scientific, I took a look at the nation's fourteen largest regions, and tried to compare their congestion and job growth levels.

In order of congestion levels, they are: 1) Los Angeles (70 hours lost to congestion per traveler), (2) Washington (62), (3) Atlanta (57), (4) Houston (56), (5) San Francisco (55), (6) Dallas (53), (7) Detroit (52), (8) Miami (47), (9) New York (44), (10) Phoenix (44), (11) Seattle (43), (12) Boston (43), (13) Chicago (41), and (14) Philadelphia (38).

If congestion was an extremely important factor in job growth, we would find that Boston, Chicago and Philadelphia are booming, and that high-congestion regions like Los Angeles, Washington and Houston were declining.

But in fact, there appears to be almost no correlation between congestion and recent job growth.   The 2000-09 job growth levels (3), in order, for these regions were: (1) Houston + 11.3%, (2) Washington + 10.6%, (3) Phoenix + 8.4%, (4) Dallas + 5.3%, (5) Miami + 4.9%, (6) Seattle + 2.5%, (7) New York + 0.6%, (8) Atlanta -0.8%, (9) Philadelphia -1%, (10) Los Angeles -3%, (11) Boston -4.3%, (12) Chicago -5.5%, (13) San Francisco -10%, (14) Detroit -20.7%. 

As you might notice, the three least congested regions all suffered negative job growth between 2000 and 2009.   By contrast, the two fastest-growing large metro areas, Houston and Washington, are also among the most congested.   On the other hand, Los Angeles and Atlanta experienced high congestion and negative job growth, while Phoenix and Seattle had relatively low congestion and moderately positive job growth. 

It could be argued that there is a lag between congestion and growth, as businesses move out of a city in response to congestion.  If this was true, prior congestion data would show a much stronger relationship to job loss than the most recent data. But in 1997, the congestion rankings were pretty similar to those of 2007.  (4) Then as now, Los Angeles was the most congested of the major urban areas, followed by Atlanta and Washington (as well as Seattle, which tied with Washington for third place).  Then as now, Philadelphia was the least congested of the major cities; the least congested regions after Philadelphia were Boston and New York.   \

In sum, the correlation between congestion and job growth is pretty weak.  It logically follows that even if the federal government reduces transportation spending, and even if such reductions do increase congestion, the overall economic effect of this result may be pretty small. 

Let me emphasize what I am not arguing: I am not arguing that congestion has no economic costs.  It makes sense to me that other factors being equal, businesses would rather locate in a place with less traffic rather than a place with more.   But I am arguing that congestion is one of many, many factors affecting our economy, and that its alleviation should be balanced against other factors. 

To put the matter another way: the Texas Transportation Institute estimates that congestion cost the national economy $87 billion per year.(5)  Sounds pretty bad, doesn't it?  But US GNP is about $14 trillion per year, so the cost of traffic congestion is less than 1 percent of GNP. 

Ironically, traffic congestion is a bit like the accommodation of bicycles.  It seems to me that Americans' quality of life would be higher if they sat in traffic less.  It also seems to me that Americans' quality of life would be higher if they could bicycle more safely and conveniently.  But I doubt that the long-term success of the U.S. economy depends on getting either issue right. 




(4) (Table 4)

(5) Id., Table 2.

Michael Lewyn is an assistant professor at Touro Law Center in Long Island.



Identifying a great, valid point

I have always thought this and I'm glad someone else mostly agrees. My argument has always been that if you already know how your commute is every day because you do it every day, you build that in your time so the "cost" is zero. It's just the unexpected delays - accidents, etc. that "cost" us as a society. Plus, when various travel modes are compared, someone assumes one mode experiences zero travel time like you can just snap your fingers and transport yourself.

There is a professor (I think UCLA) who published something that you suggested, but rejected about regions and congestion - that congestion is good because it's a indicator of strong economic activity. You might try and find his work and if you have not already.

Good post.

Food for Thought

Michael, you continue to impress with your fascination with statistical analysis.

I would have thought congestion's major factors would have been:

1) the adequacy of the roading system in the first place
2) the growth in population and regional employment since the roads were last known to be adequate.
3) the trend to centralisation or decentralisation of jobs
4) the trend to centralisation or decentralisation of residences; and/or increasing or decreasing density

This is a very complex discipline. Here is some more food for thought for you. Colin Clark would fascinate you, with his reams of statistical analysis. You should take up where he left off.

Colin Clark in "Regional and Urban Location" (1982) provides fascinating analysis of how cities naturally grow, and the exceptions. If he were still alive today, he would be one of the most valuable experts. He refers to an old proverb, "The owl of Minerva flies in the evening". That is, most wisdom is finally learnt just before it is rendered obsolete by new developments.

I think Colin Clark today, would say that urban limits and land price appreciation renders all the old "rules of thumb" obsolete. But understanding the NORMAL growth pattern of a city is helpful to identify where and why the distortions occur.

I have long understood that land prices rise from the fringe to the center, and this results in the most efficient uses of the land in the most convenient locations. It also results in urban densities being higher the nearer one gets to the center, tailing off at the point at which commercial use outbids residential.

Persons who follow Planetizen will know that I have been insisting recently, that increases in land value are in fact distorting this natural growth pattern. This is because of the shift in relationship between the value of the land and the value of the houses and buildings on it. The "cheapest option" for a home used to actually be conveniently located as well as low in price, because the structure itself, in the case of an elderly house, was worth next to nothing.

The calculation done by prospective home buyers, involved a trade off between travelling distances and cost, and also the "utility" of having a newer or older home.

But now that fringe lot prices are in the 6 figure numbers instead of 5, and the lot is as valuable as a newbuild house itself, the older homes land values alone, can be several times higher than the new fringe home. This has resulted in higher density development and infill taking place closer to the fringes. This is because it is the new "cheapest option" for households. Unfortunately for households, it is not as cheap as the historical cheapest option, AND does not have the advantage of location that the historical cheapest option did.

IMPORTANT NOTE for urban planners. Watch the "density profiles" of your metro area. If your planning is "monocentric", your average commuting distances will be increasing as your population density rises FURTHER AWAY from the center rather than closer to it.

Back to Colin Clark. Colin Clark explains that there is a further phase of growth that a city reaches if it is allowed to. Ultimately, as a city grows and its land values increase in the center, those land values eventually reach a point where they "price themselves off the market". Convenience of location dictates that they should enjoy a pricing premium, yet that premium can ultimately outstrip the economy's ability to pay the purchase or financing costs. It used to be said that the center of Tokyo was "worth" more than the entire State of California.

The result of this phase being reached, is that "nodes" other than the center itself, become more competitive and begin to "attract", like gravity, their own increased population and commercial density. The one-time center actually experiences a kind of "emptying out" at this stage. We are talking here, about a metro with many millions of population; AND low level planning and regulation. Smaller cities simply never reach this stage, at least naturally - the prices never get high enough.

My insight, I believe, is that urban limits force this phase on a metro, too early, when it is not economically viable; and monocentric assumptions persisted in, bring about yet worse distortions. The prices are forced up, closer to the center, at a rate and to a level that was only previously justified, under market forces, by very much higher levels of population and regional income.

This is the flaw in the argument that has been mounted in this forum, to the effect that if higher prices, due to location, were always a cause of densification, then higher prices still will be an even stronger force for densification. The vital point that has to be considered, is the relationship between the TOTAL VALUE OF LAND TRADED per annum, and the TOTAL GDP. When you have the value of all land pushed up tenfold (and more - this effect is still being studied, and figures of 20, 30, 60, and even 200 times, have been identified), obviously the total amount of land "traded" (and redeveloped) per annum simply has to drop, and drop considerably.

Another insight from Colin Clark: the rate at which an urban center (or "node") densifies is strongly correlated with the amount of roads provided. No urban center has achieved "mature" densities in modern, "zoned" times, based on transit systems, the roading is the vital factor. Ironically, it is population densities CONSEQUENT on the roading, that finally enable transit to be provided viably. Manhattan's transport mix of 30% transit and 70% road is exemplary.

Another insight: the "emptying out" of the urban center after it has "peaked", has its own momentum of "blight", more emptying out, more "blight", and land prices dropping. Eventually, a new metropolitan equilibrium has to be reached. One of the consequences of this central "emptying", is that the cheap land can be easily purchased by the regional government for "public space", and changes in the areas character can be cheaply influenced.

The "multi-nodal" metro thus is an entirely natural phenomenon. It brings a multitude of advantages with it, the best of which surely is the far more easily gained "location" of homes near jobs, retail, schools and amenities (especially in contrast to the monocentric model).

Of course, Colin Clark discusses geographical constraints too. The important thing to grasp is that whatever reason lies behind increases in land prices, the consequences of imbalance with GDP are the same.

I believe that it is very important to identify what "phase" any metro is in when attempting to analyse things like congestion and infrastructure planning.

Of course there are other economic factors like unions killing the goose that lays the golden egg. Surely the Detroit autoworkers who still have
jobs at $40 per hour plus perks and superannuation worth another $30 per hour, can see the connection between their "priviledge" and the death of their region.

Nathaniel Baum-Snow

It was interesting to note what this young researcher was saying too:

I think he has so far missed the role that land values play in the effects he is noting.

Pricing Land Out Of The Market

"Ultimately, as a city grows and its land values increase in the center, those land values eventually reach a point where they "price themselves off the market". Convenience of location dictates that they should enjoy a pricing premium, yet that premium can ultimately outstrip the economy's ability to pay the purchase or financing costs."

Lots of interesting ideas in this post, but this point doesn't make sense in terms of basic economic theory. Market price creates an equilibrium between supply and demand. Market price does not rise to the point where the commodity doesn't sell.

In addition, my experience in American cities is that the centers tended to be abandoned during the 1950s and 1960s, at a time when there were low land costs and rapid growth of low-density suburbs at the fringe. And centers have tended to recover during the last couple of decades, when there have been high land prices. In fact, the high land prices forced people to move to central neighborhoods which they would have avoided in earlier decades.

For example, the south Bronx was deserted and burned out by the end of the 1960s, and now it is a stable working class-neighborhood. The Uptown neighborhood of Oakland, California, was mostly vacant lots when I worked near there 15 years ago, and now it is filled with five-story condos with middle-class owners. The reason is that land for housing is so scarce in the SF Bay Area that even this scary, abandoned neighborhood has been reclaimed.

Portland, OR, has even stricter controls on suburban sprawl, and if I remember the figure correctly, hundreds of thousands of new housing units have been built in downtown. Central land wasn't priced out of the market there. On the contrary, controls on sprawl pushed new housing development toward the center.

Charles Siegel

Good points, but let's look closer.


I appreciate that you take the time to think this through and state your case.

You make 2 points that need clarification.

Under basic supply and demand theory, a commodity can't price itself at a point at which it simply does not sell at all. BUT what happens with supply and demand curves, is that the intersection occurs at different QUANTITY levels when one or the other curves shift.

With land, the "price" of all comparable land, is set by the transactions that DO occur, even though these might be few in number. This is why the "value" of ALL land can be so many times higher than GDP, and why the value of ALL land can increase so much without there being additional INCOME around to actually pay the higher prices. Sure, monetary policy does have an effect, but we are talking here about the effect on the "land price graph slope" in a metro area.

"Demand" for land will take location into account, but as prices rise in the center (as the metro area "matures"), the incentives to locate away from the center, increase. So prices rise and quantity of land traded in the expensive areas falls.

The experience you describe, of centers being abandoned, is exactly as I described from Colin Clark's research. There is a stage of "maturity" a region reaches, where new "nodes" develop that are seriously competitive with the center, and location in such nodes becomes a more viable proposition for businesses and developers than the center. There will be strong psychological expectations of high prices that make adjustments "downwards-sticky", so that the center can remain "priced off the market" until it has died altogether.

But this only occurs in the largest cities, (many millions of population) after decades of freely-allowed growth. It is VERY IMPORTANT for planners to understand just what phase of growth their city is in. Much damage is being done by planners in smaller cities that will NEVER densify naturally at the center to the extent that transit becomes viable, while land prices are pushed prematurely into the "off the market" phase by urban limits.

You are absolutely right about "dead" centers being revived because they became affordable again. This, too, is what Colin Clark said and I reiterated. Note that they revive in a different form to what they were before. Again, a lot of unnecessary expense can be caused by planners trying to impose ideas rather than just "letting it happen". But this "dead"/"blight" phase is certainly a golden opportunity for the jurisdiction to create parks at low cost, as well as conveniently located depots and infrastructure. It is an opportunity missed if they do not do so before land values rejuvenate.

It is also important that a "cheapest option" remains on the table for low income earners. This is what makes it immoral to use urban limits as a tool to force prices up. The central rejuvenation would occur anyway, just perhaps a bit slower, without urban limits pushing land prices up again. The center has advantages that do not need assistance by pushing up ALL land prices. Number one is obviously location; one would be located centrally to all the new "nodes".

I suggest too, that the number one thing that needs to be done to allow urban blight to be healed naturally, is for crime to be policed. Heather MacDonald is the expert writer on this subject that comes to mind.

Your examples of Bronx; "now a stable working class neighbourhood", and uptown Oakland, "filled with 5-story condos with middle class owners", are perfect for my thesis here. Bronx healed through policing, following which land values gradually picked up again; Oakland healed through urban limits driving prices up at a dizzying rate. The one has the working classes living as they overwhelmingly prefer, the other has driven them out altogether, condemned to lengthy commutes (if they have not left the State altogether). "Middle class" is a description dependant on perspective. The "middle classes" in cheap-land cities would not regard the Oaklanders you describe, as "one of us".

Lastly, your points about Portland. Have you looked at an urban density profile of Portland? Alain Bertaud provides them in several of his studies.
Portland, prior to Smart Growth, was a very sprawling smaller city. The central new housing development you refer to, would easily be able to occur for a little while before the changes in land values start having unintended consequences - which they have in Portland. One is, that intensity is now unnaturally high CLOSER to the urban limits. Sure, there is higher intensity where it was intended, but there is also an unintended spurt upwards in the density curve, towards the urban limits.

I believe that this will become a common trend, and there are thesis papers waiting to be done on this. I watch keenly for this effect everywhere I go, where I know urban limits are in force and median multiple prices are high. Try it for yourself. You will see "infill" development and large houses and lots being split into 2 or 3, way out near the urban limits; and when the houses already there are not all that old, too. Meanwhile, closer to the urban center, there will be a whole lot of old houses on large sections, with a few apartment blocks and condos sprinkled among them - and no new ones in prospect; and almost none of the infill and cutting up of houses and lots that you saw further out.

The consequence of this is that average commuting distances, under the monocentric model being pursued in these cases, actually is being driven UP, not down.

The other effect demonstrated in Portland, is demonstrated by Portland economist Randall Pozdena in the paper "The New Segregation". Lower income groups have been driven out of Portland altogether over the last couple of decades. Also, look at the effect on the rate at which subprime mortgages occur. The question has to be asked, what happens if planners succeed in eliminating ALL affordable property by having limits imposed on all the jurisdictions that currently do not have them?

At least some congestion is necessary

Congestion is a fact of life for places good and bad, excepting maybe those wastelands beyond redemption. If a place has any vitality it is likely to have a degree of congestion. Of course a great deal of places experience congestion without being all that vital due to poor planning and design. So being bright and self-interested, we Americans will exhaust every last bit of our national wealth in a futile attempt to keep our driving clear and free. We separate land-uses, require free parking as a birthright, keep density as low as possible, remove superfluous sidewalks, and eliminate hazards to drivers such as trees and bicyclists. Yet we still find ourselves perpetually annoyed with the "congestion" just about everywhere anyone lives.

Some of us of course have accepted reality. Some of us have found that living in communities where there are transportation choices makes a big difference. Sure getting to work on the train may take just as much time as driving. BUT, it's loads cheaper and provides extra time to read, listen to music, do cross-words, or just space-out. Plus, by living in walkable neighborhood, not everything requires a train trip like going to work might. Need a gallon of milk? Want to grab a coffee or a bite to eat? Getting some fresh air in the park? All within a couple minute's walk from home, in a well-planned, well-built place. So yes, driving anywhere you can expect traffic, but some places have a lot better options if you want to get out of traffic and enjoy where you are...

Balanced Approach to Transporation is Important

A cogent and thought-provoking blog. Congestion happens naturally in areas with strong economic activity -- even those that take a holistic approach to developing a broad transportation infrastructure. But this view should not allow us to get complacent about the need to continually expand and maintain our nationwide transportation networks to serve growing populations. Beyond city congestion, the larger issue of nationwide transportation infrastructure is critical. One reason that many third-world countries and even emerging economies still face economic difficulties is that goods and services are not easily and effectively distributed to all areas or between cities due to sub-standard or insufficient transportation infrastructure in comparison to their populations.

A balanced and careful prioritization of transportation system funding is important. But urban congestion is just one symptom that the USA's transportation infrastructure is under severe stress: for the simple reason that much of it has not been upgraded or expanded since it was built; has exceeded its intended lifespan; and now supports populations that are far larger than predicted when they were constructed. The U.S. population now tops 300 million -- up from 200 million just 45 years ago. It's expected to hit 400 million in 40 years. Here in California, the population has doubled to 38 million in just 40 years. But investment in existing infrastructure is lacking, as are planning and construction of the infrastructure needed to support the future population. In that time, Interstate 5 between LA and the SF Bay Area -- the most vital transportation artery in the state -- has remained at 4 lanes with significant reductions in average speeds due to increasing congestion. The benefits that the highway provided when first completed have now largely evaporated. No widening is planned, and high-speed rail could be decades away.

Public health, a strong economy, and a high quality of life demand a robust transportation infrastructure. Heavily congested Europe provides a good model: with top-notch, high-quality infrastructure for cars, abundant mass transit and (in many cities) dedicated networks for bicycles. These are not mutually exclusive needs and they ALL require tending to. They won't end congestion, but the demonstrate that Europe is not complacent about planning holistically for its long-term transportation needs.

Paul Davis

Prepare for the AICP* Exam

Join the thousands of students who have utilized the Planetizen AICP* Exam Preparation Class to prepare for the American Planning Association's AICP* exam.
Starting at $245

Essential Readings in Urban Planning

Planning on taking the AICP* Exam? Register for Planetizen's AICP * Exam Preparation Course to save $25.

Get the "Green Bible" of planning books

Understand the complexities of planning at the local level while preparing for the AICP* exam.
Habitat blockitecture tower

Build the world you want to see

Irresistible block set for adults when placed on a coffee table or desk, and great fun for kids.