Austin Contrarian

I forgot about Cap Metro's interests

Last week, I concluded that the plan to add two tolled lanes to MoPac would make everyone better off than either maintaining the status quo or adding two free lanes.  But it occurred to me that I haven't considered everyone's interests.   Cap Metro's, specifically.  Isn't it likely that Cap Metro will be worse off?

As I noted last time, congestion-pricing will be a boon to bus riders.  They will get much shorter, more predictable rides.  But these buses run from the stations (Leander and Lakeline) that provide the bulk of Cap Metro's morning boardings.  Cap Metro is to some extent competing against its own buses for its rail riders.   (Cap Metro's surge in rail ridership last year occurred after it scaled back its express bus service from Leander and Lakeline.  The reduction in bus service might not have been the sole cause of the bump in ridership, but it surely played a part.)  

The shorter, more reliable bus trips will inevitably attract some current rail riders.  What's more, Cap Metro will be under a lot of pressure, I predict, to expand express bus service when the tolled lanes open.  The benefits for express bus riders are one of the big selling points of congestion pricing.  Free-flowing lanes will make it possible to offer reliable express bus service during the very the peak of the peak of the morning commute.  Cap Metro will be in the difficult position of either providing great, new express bus service at the cost of its rail ridership or refusing to provide great, new express bus service to preserve its rail ridership.

Perhaps it's not useful to frame this in terms of Cap Metro's interests.  Cap Metro's mission is to provide quality, cost-effective public transit.  If changes occur to the transportation network that allow Cap Metro to provide better public transportation, then we should count that as a win, even if it puts Cap Metro in a difficult position institutionally. 

So perhaps this is a better way to frame the point:  adding congestion-priced lanes to MoPac will put more pressure on the Red Line.  Cap Metro is running the Red Line today at an unsustainable subsidy of $33.98 per rider.  I don't know how Cap Metro plans to reduce that subsidy to a reasonable level.  I'm pretty sure, though, that improved bus service on MoPac will make the path more difficult. 

The Freeway Congestion Paradox

I've gotten some pushback on my claim that adding two congestion-priced lanes to MoPac will increase capacity more than adding two free lanes.

This is the Freeway Congestion Paradox:

Chen and Varaiya, in their article entitled "The Freeway Congestion Paradox," have demonstrated that, once freeway vehicle density (measured in vehicles per mile) exceeds a certain critical number, both vehicle speed and vehicle flow (measured in vehicles per hour) drop precipitously.  They have demonstrated the phonemenon with actual data from a section of westbound I-10 in Los Angeles.  Until 5:10 am, a flow of 2,100 vehicles per lane per hour is maintained, at a speed of 58 mph.  As density increases after 5:10 am, speed steadily drops, until at 7:00 am speed is a stop-and-go 15 mph, and flow decreases to 1,300 vehicles per lane per hour.

The Federal Highway Administration commissioned a study in 2008 that looked at data from Washington, DC freeways.  From the summary:

Highway studies have determined that once traffic volumes exceed the capacity of the roadway, the system can rapidly "break down" to the point where all traffic slows markedly, and the capacity and throughput of the roadway drops precipitously.

They produced neat graphs depicting the drop in capacity caused by stop-and-go traffic, like this one:

This chart pairs vehicle speed and total volume carried (see pp. 15-16 of the report for more discussion).  The dots are connected in temporal sequence, so it's possible to follow the evolution of the traffic jam.  The graph has a a random-walk component, but basically it confirms two things.  First, low throughput volumes are associated with low speed (i.e., stop-and-go traffic).  Second,  after the critical threshhold is passsed, it takes traffic a long time to recover. 

Here is the same chart, except I have drawn over some of the line segments in red.  This shows that once the tipping point is reached, traffic speeds and volumes drop and take a long time to recover.   (Just trace the red line segments starting at "A").

Congestion pricing would have prevented congestion from reaching the tipping point, effectively increasing the highway's capacity.

Rush-hour traffic on MoPac today is the stop-and-go 15 mph type.   Adding a new, free lane would probably shorten the period when the highway is clogged, but it would still be clogged for long stretches of time.  During these periods, a new, free lane would carry less traffic than a tolled lane.

When congestion is not bad, the toll should be low.  That's one of the attractive features of variable congestion pricing -- the price, in fact, should be zero when the marginal cost of congestion is zero.  (According to the Statesman, CTRMA might charge a nominal amount like a quarter to avoid confusing people.  Which confuses me.)  Put differently, the new lane should "act" like a free lane when traffic is not bad.

Congestion, envy and equity

The Central Texas Regional Mobility Authority plans to add two lanes to MoPac between Lady Bird Lake and Parmer Lane.  The cool part of the plan is that it intends to use dynamic congestion pricing -- i.e., the toll will fluctuate as necessary to keep traffic in the the tolled lanes flowing at 50 mph.   The six existing lanes will remain free.

This is how we ought to add new capacity.  It will make everyone better off.   The people who choose to pay the toll will be better off because they value the time savings more than the cost of the toll.  Bus commuters will be better off -- they might be the biggest beneficiaries, in fact -- because they will get a suddenly much shorter commute for (I presume) the same bus fare.  Drivers who continue to use the free lanes will endure slightly less congestion, even if it's just a narrower period of peak congestion.  Finally, taxpayers, if not better off, will be no worse off because they won't have to pay for the extra capacity.  The capacity will be paid by those who value and use it.

Everyone will not only better off compared to the status quo, but . . . and this is the key point . . . they'll be better off than if the two new lanes were free.   Adding two free lanes would reduce congestion, too, but they would reduce congestion less than two new tolled lanes.  Tolling congested lanes increases their capacity.  A congestion-priced lane can handle 1,800-2,000 cars per hour; an unpriced lane during peak congestion will handle less than half of that.  Somewhat counterintuitively, perhaps, charging for a congested lane is a sure way to get more cars through it.

Obviously, drivers happy to pay the tolls to skip congestion would be worse off if the new lanes were free rather than tolled.  Bus riders who would lose out on a faster commute would be worse off if the new lanes were free.  Taxpayers stuck paying for capacity that they might not value would be worse off if the new lanes were free.   But, because tolling lanes increases their capacity, even the drivers who would not use the toll roads anyway would be worse off if the new lanes were free.  They would have to contend with more congestion.   

Congestion pricing is a no-brainer.  So I was puzzled when Ben Wear raised the equity angle in a companion column.   After pointing out that all drivers would benefit from the congestion reduction from the new lanes, he concluded:

Undoubtedly however, those with money will be able to use the variable-price lanes more often than those on a budget.

It may become something of a spectator sport for those caught in MoPac congestion to count the number of Lexuses, Mercedeses and Escalades that fly by in the express lane.

It ain't necessarily fair. But in the words of John F. Kennedy (who certainly could have afforded to use this lane), life is unfair. And it will help pass the time.

It's strange to complain about fairness when everyone will be better off with new tolled lanes than new free lanes.  

But it's especially strange to complain about fairness because, once the lanes open, the only people on MoPac who will be causing a net social loss will be the drivers in the free lanes.

 This is basic congestion theory but most people still don't get it.  When you enter a congested highway, your mere presence inflicts a cost on everyone else.  Every car behind you is delayed just a little bit, maybe only a few seconds each, but the delay affects a bunch of cars.  Getting in everyone else's way imposes a social cost.

There's not a lot a driver can do about that today.  We can't even say for sure what cost each driver is imposing on everyone else.  But once the toll lanes are built, we'll know.  We'll know because the congestion cost each driver imposes on his fellow drivers will be approximately equal to the price of driving in the congestion-free lane.  That's one of the quirks of the economics of congestion -- it turns out that the cost of congestion a driver imposes is equal to the congestion-clearing price.

Once the toll lanes are built, drivers will have a choice.  The people who enter the tolled lane will be precisely the drivers willing to pay the cost of their congestion.  The toll lane will allow them to pay that cost and get out of the way of everyone else.

The people who don't pay the cost are people for whom, say, 15 minutes in time savings isn't worth the $3 toll or whatever it turns out to be.  That's most people, presumably.  But saving 15 minutes is worth something to them.  If you value the time savings at $1, you would incur a $2 loss by taking the tolled lane.  But you will impose a $3 loss on your fellow drivers if you stay in the free lanes, because we know the cost of your congestion is equal to the toll.  Staying in the free lanes produces a net social loss of $1.

Congestion is socially-sanctioned selfishness.  It's ok only because everyone does it.  But if we'd get used to thinking in terms of costs we're inflicting on everyone else, there'd be less moaning about those paying for the cost they're imposing while we mooch our way down the free lanes.  

Texas is studying a tax on the number of miles you drive

News to me:  Texas will study a tax on miles driven:

If you don't like gasoline taxes, here's an alternative: a tax on the number of miles you drive in a year.

The Texas Transportation Commission has directed a fresh study of the idea, and it is not alone.  There are pilot projects in other states and nationally to gauge how such a tax would work.

Texas transportation officials say the study is meant to help give lawmakers information on options ahead of their next regular session in 2011, when they confront a funding squeeze that is expected to drain the highway fund of money for new construction contracts by 2012.  “We need to think differently about how we fund transportation,” Texas Transportation Commission Chairwoman Deirdre Delisi said at a Texas Taxpayers and Research Association forum in November.

Delisi said the vehicle-miles-traveled tax idea is controversial, but should be discussed because revenue from the state's main source of transportation funding, the motor fuels tax, is declining.  The gasoline tax has not been raised since 1991.

The commission asked the Texas Transportation Institute, which is part of the Texas A&M University System, to take the lead on the study.  Commissioner Fred Underwood has emphasized that the commission's goal is to give lawmakers alternatives.

There are good arguments that a gasoline tax is a more efficient tax on externalities.  Heavier vehicles do more damage to roads; heavier vehicles use more gasoline.  Emissions vary with the amount of gasoline used (ignoring vehicle technology).  A miles-traveled tax perhaps better reflects a driver's accident risk to others, but larger vehicles impose a higher risk on other drivers, bicyclists and pedestrians than smaller vehicles.  (Neither accounts for congestion externalities.)

But two drivers who "consume" the same number of miles enjoy equal benefits.  It's fair in that sense to require them to pay the same amount.   A gas tax and miles-traveled tax thus serve different purposes:  a gas tax is a tax on externalities; a miles-traveled tax is a tax on consumer surplus.

"No road we built in Texas paid for itself."

From Streetsblog Capitol Hill:

Over the past two days at the Congress for the New Urbanism Project for Transportation Reform conference, attendees have called for reform at local, regional, and national levels. In a panel debate about the future of transportation funding and the role of regional planning through MPOs, several speakers argued that the foundation of transportation and development funding had to be systematically overhauled.

Mike Krusee, [AC: former] chairman of the Texas House of Representatives Transportation Committee, said that financial problems were more significant than environmental, though they should be tied together in the same discussion.

"The reason there's not a new transportation bill is because there is no money. We've hit the wall of unsustainability on how we finance the transportation system," he said.

Krusee asserted it was urgent and necessary to understand the nature of this broken financial apparatus and to develop solutions to fix it. In Texas, he said that, on average, it cost the state 20-30 cents per person per mile to build and maintain a road to the suburbs, yet drivers only pay on average 2-3 cents per mile through the gas tax, vehicles fees, etc.

"What we found was that no road that we built in Texas paid for itself," said Krusee. "None."

The expense to build roads and utilities further and further from the urban cores not only drove costs to unsustainable levels, it created an imbalance in who paid for growth. Over the past 50 years, Krusee argued, the federal government used tax money that came by and large from cities to subsidize roads to areas without access otherwise.

"City dwellers have subsidized the land purchases and the development costs out in the suburbs," said Krusee. What's more, the gas tax, which city dwellers pay when driving on city roads, but which goes to freeways largely outside of urban cores, is "a huge transfer of wealth from the cities to the suburbs to build these rings."

Krusee said building the interstate system was initially a good thing, because if facilitated interstate commerce and increased the productivity of cities. Now however, because of congestion caused by ever longer commute patterns, system productivity is in peril. "What's happened is the federal government has basically reneged on the deal. By subsidizing highways out to the suburbs, it's no longer efficient for truck traffic, for goods and services and people to move between cities in the United States because those roads have been hijacked by all the commuters."

Krusee, by the way, represented suburban Williamson County.

Krusee's assessment matches TxDOT's own internal assessment.  (This actually should be no surprise since Krusee's committee relied on TxDOT for data.)   TxDOT, for example, concluded that the 15 miles of SH 99 from I-10 to US 290 will cost $1 billion to build and maintain over its lifetime, while only generating $162 million in gas taxes -- just 16% of the total cost.

Some of us get swept up in the rhetoric sometimes, but roads aren't unmitigated evils.   Obviously, we need roads.  Just as obviously, I think, we will continue to need new roads.   But new roads should be built only where drivers are willing to pay for the new capacity.   And the only way to gauge that demand is to price existing roads properly; the revenue they generate will tell us when it is time to add to add that capacity.

No

I agree with Ryan Avent that this WSJ piece makes a bizarre argument against congestion pricing:

By requiring car drivers to pay a fee to drive in a city at peak hours, congestion pricing reduces traffic and raises money that can be used to support public transit—both worthy goals.

Yet congestion pricing has dubious environmental value. Traffic jams, if they're managed well, can actually be good for the environment. They maintain a level of frustration that turns drivers into subway riders or pedestrians.

The author contends that by improving traffic flow, congestion pricing induces more driving, which in turn increases environmental damage:  "If reducing it merely makes life easier for those who drive, then the improved traffic flow can actually increase the environmental damage done by cars, by raising overall traffic volume, encouraging sprawl and long car commutes."

Ryan makes the right arguments.  I think Owen's got the economics flat wrong, which I'll elaborate in another entry.  He's wrong because congestion pricing does not induce more traffic and because congestion pricing encourages more compact cities rather than sprawl.

But Owen's also got the environmental impact wrong.  He ignores that traffic jams impose a cost -- a significant cost -- not imposed by free-flow traffic.  That's health.  Traffic jams concentrate particulate emissions in one location.  High levels of particulate emissions are bad for people.  They cause all kinds of health problems -- asthma and other respiratory illness, obviously -- but also higher risks of infant mortality.

If you need evidence, read this spanking new paper (h/t Matthew Kahn).  It compares infant mortality rates near toll plazas to infant mortality rates after the toll plazas were replaced with an EZ pass system.  They found that the EZ pass system "reduced the incidence of prematurity and low birth weight among mothers within 2km of a toll plaza by 10.8% and 11.8% respectively."  Because toll plazas are notoriously congested and highways with an EZ pass system are not, this is a good proxy for the health effects of traffic bottlenecks.

Environmentalism is not only about carbon emissions.  The wellspring of the movement was Rachel Carson's crusade against environmental contaminants that (allegedly) made people sick.  Any calculation of environmental cost that omits the costs of making people sick is not just a bad cost-benefit analysis, it's bad environmentalism.

Back and forth

It never surprises me to hear suburban commuters bash congestion tolls.  But I am surprised when a sophisticated transit blogger bashes them.  Yonah Freemark, who writes the very good the transport politic, is the guilty party.  He and Ryan Avent have been going back and forth and back again.  

Yonah's objection is equity:

A huge percentage of the U.S. population pays far too much for transportation; to put it simply, most working adults have no choice other than to own a vehicle and often to drive it dozens of miles every day. Making driving more expensive is a great way to devastate the already impoverished.

It’s true, tolling highways would save “money, time, lives, and emissions.” But it would also sacrifice the mobility of a large segment of America, because the reduced congestion would be a result of the poor and the middle class choosing not to drive because of expense, not because of choices made by the wealthy.

He argues that our public transit systems are too underdeveloped in most places to cure this inequity.

Ryan responds with several good points, but I want to elaborate on why the regressivity argument leaves me cold.

First, congestion tolls are less regressive than toll opponents claims.   The low-income do pay tolls, which means they value the time savings more than the price of the toll.   Plenty of low-income drivers take the toll lanes on SR 91 in California, for example.  The reason many do is that they have little flexibility in their schedules.  They have to be at work or at the daycare center at a specific time.  To compensate for congestion, they must leave much earlier to guarantee a timely arrival.  They thus suffer two types of costs:  scheduling delays and travel delays.  (The value of scheduling delays is very hard to measure, which means that studies like the TTI report almost surely understate the cost of congestion.)  Affluent professionals often have more flexibility.

Second, tolls encourage a number of shifts.  Yes, shifts to transit, which seems to be Yonah's main concern, at least when the transit system is underdeveloped.  But they encourage other shifts, too.  Shifts to other routes and shifts to other times.   Commuters are the least likely to be nudged to other routes or times.  The most sensitive are those who use congested roads for local trips.  Take the soccer mom who hops in the SUV and enters a congested highway to get to the grocery store a mile down the road.  She imposes enormous costs on others.  Tolls make her internalize those costs and nudge her to use the local streets. 

Perhaps this is regressive.  But it doesn't evoke much sympathy from me.  A congestion toll is a charge for getting in everyone else's way.   Behind every claim of regressivity is the assumption that drivers are entitled to get in everyone else's way even when it is not worth all that much to them. 

And that leads to my last point.  The regressivity argument wouldn't move me even if it were true. This is one of those cases in which our desire for efficiency should trump our concern with regressivity.

Here's an apt analogy.  An amusement park owner decides to throw open the gate to all comers on a first-come, first-serve.  Naturally, a long line forms.  But this isn't a typical line where newcomers go to the back of the line.  No, in this queue, newcomers elbow their way to the front of the line and force everyone behind them to wait a little longer.  

Naturally, chaos ensues.  No one leaves for the park knowing how long it will take to get in.   There is "queue rage" and general aggravation.  The park owner loses business.  And many low-income parents find themselves worse off because they have a smaller window of free time and dislike the chaos and aggravation as much as anyone else.  (I've always thought it patronizing to assume that low-income drivers put such a low value on their time and aggravation.)

Now, in the real world, the queue would never work this way.  Those at the rear of the line would use informal sanctions (fist fights) to deter queue jumpers.  And, in fact, there would be only a short  line to get into the park in the first place because the owner would charge for admission.

But this is exactly the crazy system we use to ration highway access.  For whatever reason, our cultural norms have evolved to tolerate a free-for-all.  In almost every other situation we ration scarce goods using price.  Sure, that's regressive in the sense that the poor have less money to spend on things.  But we tolerate some regressivity elsewhere because we recognize that (1) the gains in efficiency outweigh the equity concerns; and (2) there are better ways to ensure an egalitarian distribution of wealth than creating artificial shortages, chaos and mayhem.

Regressivity is not the be all and end all.

Cross-posted at Urban Returns.

TTI hyperbole

There is no question that congestion is a serious problem in this country -- and a crippling problem in some cities.  Nor is there any question that congestion is much worse than it was ten years ago, even if high gas prices and the recession have caused it to level off in some places.

But the Texas Transportation Institute is guilty of a bit of hyperbole when it announces that "[t]he overall cost [of congestion](based on wasted fuel and lost productivity) reached $87.2 billion in 2007 – more than $750 for every U.S. traveler."

TTI uses free-flow traffic as its baseline for measuring congestion.  In particular, it defines "Delay per Peak Traveler" to be "[t]he extra time spent traveling at congested speeds rather than free-flow speeds divided by the number of persons making a trip during the peak period." (Summary report p.1)

Honestly, free-flow speed is not the right baseline.  There is no realistic, hypothetical state of the world in which we would experience perfect, free-flow traffic everywhere.   It would not be feasible to build enough roads (or charge enough for them), particularly since free-flow speeds would entice more drivers onto the road.   So to imply that there is $87 billion of waste to be saved -- and I think TTI does imply this -- is simply wrong.  The TTI report makes the news every year thanks to this spectacular estimate, but this one probably belongs on the tabloid pages.

Anthony Downs makes this point in Still Stuck in Traffic, pp. 23-24 (available on-line at Google Books).   He makes several other reasonable criticisms of TTI's methodology.   Hedoes recognize that  TTI's methodology and findings are useful, and he does not dispute (nor do I) that congestion is a serious problem that we can't simply build our way out of.

(Edited and cross-posted from  Urban Returns.)

Density and congestion, etc.

Whenever I see new data on cities, I'm always tempted to match them to the cities' weighted densities, if for no other reason than no one else does it.  And so with the Texas Transportation Institute's latest report on city congestion.  TTI found a wide range in hours lost due to congestion per year -- e.g., in 2007, Lost Angeles drivers lost an average of 70 hours per year to congestion;  Cleveland drivers, just 12.  (The relevant TTI table is here  (pdf).)  Does weighted density partly explain this variation?

The answer is "No," based on my admittedly simplistic analysis.  

Below the jump I have three charts plotting, for 33 cities, weighted density, standard density and total population against hours lost per traveler to congestion.   (The 33 cities include the 31 largest urbanized areas -- excluding New York City, which is always an outlier -- and Austin and Honolulu.) 

As the scatter plot shows, weighted density explains virtually none of the variation in congestion (adjusted R² = .06).   Standard density explains a bit more  (adjusted R² = .19).  And total population, a bit more than standard density  (adjusted R² = .28). 

Punching above its class

The Texas Transportation Institute has issued its annual report on congestion in American cities.  As usual, Austin doesn't do well, punching well above its class.  Tthe average Austin traveler experienced an annual delay of 39 hours in 2007.    That's up from 32 hours in 1997, although no worse than 2006.   

Ryan Avent puts the Austin figures in perspective:

The average traveler in the New York metro area faces 44 hours wasted per year, for instance, while the average traveler in Los Angeles loses 70 hours per year to congestion, even though New York’s metropolitan population is much, much larger than LA’s.  More interesting still, Austin and Raleigh aren’t that far behind New York with 39 and 34 hours wasted annually, despite the fact that both metro areas have less than two million people while greater New York is home to 20 million people.

TTI also calculates the percent of peak period travel that is congested.   Austin again fares poorly -- its roads are congested for 70% of the peak travel period: 

By comparison, Dallas's roads are congested for only 66% of the peak travel period.  Houston is marginally worse at 73%.   Austin's peers include New York (69%), Detroit (71%) and Baltimore (69%).  

One has to wonder what damage Austin's congestion has wrought to downtown.   Businesses want to locate where their employees can get to them, and high peak-period congestion means that commuters must fight traffic to get to work most of the time.   I suspect Austin's downtown suffers more heavily from peak-period congestion than any other employment center in the area since MoPac and I-35 -- downtown's principal commuter routes  -- are the two most congested roads in the area.     Eventually we must recognize that our congestion will stunt downtown's growth, if it hasn't already.   We wil lose the economies of scale and agglomeration benefits that a vibrant downtown would  yield.

You know, of course, that there is only one solution.