Ten homes is not enough

This past week, the IJ trumpeted 10 new affordable homes built in Novato, calling it “proof that Marin has room for affordable housing.” [1] These homes, which took 5 years to build, are undoubtedly welcome for their new owners, but the editorial makes a mockery of Marin’s housing crisis and the depth of the county’s need.

Over the past 7 years, Marin added about 785 homes – an increase of about 0.7 percent. [2] This might be impressive if it weren’t far outstripped by population growth of 4.4 percent, [3] six times faster than housing stock, or even more outstripped by jobs growth of 12.2 percent, [4] 17 times faster than housing stock.

Image by the author.

Image by the author.

In that context, a 10-home project is impressive mostly because it increases Marin’s pitifully meager annual housing production by 10 percent, but it is a drop in the bucket compared to the crisis-level shortfall Marin is facing today.

If housing construction kept pace with job growth – never mind the regional housing need – the county would have added over 13,000 new homes, almost 2,000 per year. With inclusionary zoning, that would mean 400 affordable homes every year, not a mere 10 every 5 years.

This is a pace of construction that Marin is unlikely to ever meet, but it shows the sheer size of the hole we’re in. Marin doesn’t just need a few more homes; it either needs to increase its construction pace by an order of magnitude (or take active steps to hurt its economy and stop the creation of new jobs).

Celebrating the opening of 10 homes is great, and the work done by Habitat for Humanity is bold. But until we get to where this sort of opening is small potatoes, it’s like celebrating your D-average: nice, but also kind of sad.

Works cited

[1] Marin Independent Journal, “Proof That Marin Has Room for Affordable Housing,” Marin Independent Journal, July 31, 2017, sec. Opinion.

[2] US Census Bureau, “Building Permits Survey” (Washington, DC: US Census Bureau), accessed August 6, 2017.

[3] Ibid.

[4] Center for Economic Studies, “Quarterly Workforce Indicators” (Washington, DC: US Census Bureau), accessed August 6, 2017.

Planning as physical philosophy

The Papers Series are lightly edited papers I wrote for classes at Cornell University as I studied for a Master's in Regional Planning. Often, references will be to printed books rather than to websites. If you want to check it out yourself, contact me and I will see if I have a scanned copy available.

This paper was originally written at the end of my first semester. It details the evolution of my understanding of planning as a discipline, not just as a subject area, through the lens of the semester's readings.

Since I began the program at Cornell, my thoughts and ideas about planning and planners have changed markedly, though I would classify the change as evolutionary rather than revolutionary. My once-clear perspective as a Glaeserian urbanist has become clouded by notions of justice, equity, participation, democracy, and humanity’s aptitude to tear down each of those. That said, the cloudiness in my outlook is one of added complexity, not of rejection. Like salt dissolving into water, I suspect these notions will dissolve into my urbanism, leaving something that looks similar to the original product but is quite different in taste and use.

My background as a blogger and advocate in Washington, DC, and Marin County steeped me in the world of activism, which stretches beyond the world of traditional planning and into engineering and politics. Alon Levy divided a subset of these activists, transit activists, into what he termed technicals and politicals:

Politicals are the people who tend to trust the transit authorities, support a general expansion of all rail transit projects, and believe the primary problem is defeating oil-funded anti-transit lobbies. Technicals are the people who tend to distrust what the authorities say, and prefer their own analysis or that of technically-minded activists; they support transit but are skeptical about many projects, and treat agency inertia and turf wars as the primary obstacles for transit revival. [1]

While Levy was concerned with transit activists only, the distinction is usefully applied to planning activists as a whole. In my writing I have striven to be technically minded first, leading me to believe that poor planning outcomes were fundamentally problems of rules and subsidies rather than inherent injustice. These were problems of “agency inertia” rather than unspoken bias or injustice, and I wrote from this perspective. [2a, 2b, 2c, 2d] As I anticipated Cornell, I thought that being able to easily discourse on FAR, traffic flow, zoning, bike lanes, rail procurement, and a host of other topics that I believed were core to planning would prepare me for class. I was wrong.

The great theme of my first semester was the exploration of justice and equity. From Samuel P. Hays [3] to June Manning Thomas [4] and a huge body of literature between and beyond, planning has roots that extend past the physical and deep into the philosophical. Planners are not simply engineers who must play politics to be allowed to do quality work, as Levy’s technicals might want. No, planning arises from the very political concerns of justice. Any answer to the question, “How ought we live?” must necessarily include economics and planning - the "we" in that question - as well as personal ethics and morality.

Though, as a planning course, we did not explore the question of how we ought to live directly, we did touch on the effects whatever answer must have on planning. In short, these principles are:

  • The promotion of economic equality
  • The promotion of equitable power distribution
  • The recognition and celebration of diversity in all forms
  • Reflection upon one’s own failures and successes

None of these planning principles deal with physical planning. Instead, planners must place those physical aspects of the profession in a subordinate position to justice. To this technically-minded armchair planner/blogger, this inversion of priorities within the planning profession was a surprise, though in retrospect it ought not to have been.

My undergraduate education in politics was steeped in concerns of justice even though contemporary politics rarely addresses justice directly. Rather, justice must be worked out through public policies. As a field with ties to many public policy arenas, planning must also look to justice as a guiding light. Curiously, these principles were outlined in papers and books long before the concept of justice itself was analyzed by planners [5a, 5b]. The success of Susan Fainstein’s The Just City [6] and Leonie Sandercock’s Towards Cosmopolis [7] strongly implies that there was a yearning for planning-focused philosophical examinations of the nature of justice, but the paucity of work on the subject prior to Cosmopolis implies that the cart came before the horse.

Although not as surprising as the importance of justice to planning, the prominence of politically left-wing solutions to myriad planning problems was unexpected. Bill Goldsmith’s twin lectures [8] on the progressive city, for example, were stridently and proudly leftist. [Note: the lectures were class lectures and are unavailable. However, Goldsmith's book, Saving Our Cities, was the source for much of series and is therefore recommended.] Dolores Hayden’s discussion of gender and city planning led to a decidedly collectivist model [9]. Sandercock praises the activist planner who leaves her career behind to fight for the underclass [10]. Meanwhile, Smart Growth activists (for example James Bacon [11], Streetfilms [12], and David Schaengold [13]) have actively tried to appeal to an economically conservative audience that would find such analyses repugnant.

The contrast between these two thrusts of planning thought is striking. Social justice is firmly based in progressive and left-wing theory, while prosaic discussions of urban finance find their basis in classical economics and right-wing theory. Although I expected some emphasis upon the former rather than the latter, I did not expect the political left to so dominate the field.

I wish to bring a variety of things forward with me into my planning practice, but perhaps nothing is more important than the impossibility of making people good. Although hardly a new concept, the at times utopian, at times downtrodden perspectives of Marxist and neoliberal urbanists alike remind me of the folly of trusting people to do the right thing. Neither Marxism nor neoliberalism offer systems that get people to actually abandon injustice. Rather, they both seek ways to restrain the fundamentally flawed nature of humanity.

In a planning practice, then, we must always be aware of the infinite creativity of people to seek their own well-being at the expense of others. No matter how great our plan, someone will probably find a way to muck it up. Since we must also be optimistic, as we planners ought expect our plans to be followed to some degree, our aim ought to be to be clear-eyed optimists who will expect failure to lurk around every corner.

When failure or success do come, self-reflection as described by Raphaël Fischler [14] must be at the core of my practice. Why did that meeting work and this one did not? What has gone right with that project after 5, 10, or 15 years and what has gone wrong? Did I bake failure into the project, or were outside forces at work? How might I inoculate my projects against such outside forces in the future?

Self-reflection must also mean a constant search for a truer meaning of justice. Justice, the context of planning, goes beyond one's personal definition of what is just and must respect for others’ definitions of justice. If I allow my own concepts of justice to fossilize, I risk losing some of my capacity to respect and honor diverse ideologies. Given that planners are also called to a more equitable power distribution, I must especially approach disempowered communities with humility. Self-righteousness is antithetical to the aim of both the former aim (respecting and honoring ideologies other than my own) and the latter (empowerment of the disempowered).

The first semester of CRP was an unexpected and delightful surprise. It was far more introspective, theoretical, and exciting than I had ever imagined it to be. Planning itself is clearly quite bold. As a field, it challenges how we order our society and asks us to truly be broad-minded in a way that pure political science does not. A politician must eventually find a constituency and find how to work justice within the confines that constituency sets, but a planner is called to form a more just society beyond whatever constituency she may be most comfortable with. I am happy to have joined that calling.

Works Cited

[1] “Politicals vs. Technicals: The Primary Division of Transit Activists,” Pedestrian Observations, June 28, 2011.

[2] David Edmondson, “A Greater Marin,” The Greater Marin, March 19, 2012; David Edmondson, “Reducing Passenger Train Procurement Costs” (Washington, DC: Competitive Enterprise Institute, June 5, 2013); David Edmondson, “Being Marin Again,” The Greater Marin, December 9, 2013; David Edmondson, “Tautological Housing Study Reminds Us That Demand Is More than Skin Deep,” The Greater Marin, February 11, 2015.

[3] “The Politics of Reform in Municipal Government in the Progressive Era,” The Pacific Northwest Quarterly 55, no. 4 (1964): 157–69.

[4] “Social Justice as Responsible Practice: Influence of Race, Ethnicity, and the Civil Rights Era,” in Planning Ideas That Matter: Livability, Territoriality, Governance, and Reflective Practice, ed. Bishwapriya Sanyal, Lawrence J. Vale, and Christina D. Rosan (Cambridge, MA: The MIT Press, 2012).

[5] John H. Mollenkopf, The Contested City (Princeton: Princeton University Press, 1983); Doug McAdam, Freedom Summer (New York: Oxford University Press, 1990).

[6] The Just City (Ithaca, NY: Cornell University Press, 2011).

[7] Towards Cosmopolis: Planning for Multicultural Cities (Chichester, England: Academy Press, 1997).

[8] “The American City Today” (Lecture, Introduction to Planning Practice and History, Cornell University, October 7, 2015). See Saving Our Cities: A Progressive Plan to Transform Urban America (Ithaca, NY: Cornell University Press, 2016).

[9] “What Would a Non-Sexist City Be Like? Speculations on Housing, Urban Design, and Human Work,” Signs, Supplement on Women and the American City, 5, no. 3 (1980): S170–87.

[10] Towards Cosmopolis.

[11] “Too Little Density, Too Much Road Surface,” Bacon’s Rebellion, November 12, 2015.

[12] Elizabeth Press, William Lind: A Conservative Voice For Public Transportation (New York: Streetfilms, 2009).

[13] “Why Conservatives Should Care About Transit,” Public Discourse, April 17, 2009.

[14] “The Reflective Practitioner,” in Planning Ideas That Matter: Livability, Territoriality, Governance, and Reflective Practice, ed. Bishwapriya Sanyal, Lawrence J. Vale, and Christina D. Rosan (Cambridge, MA: The MIT Press, 2012).


Valley to Valley: linking SMART and regional rail

The dream of urbanites across the central Bay Area, as they gaze out towards the golden hills of the Marin Headlands, is BART. They share that dream with the suburbanites who stare back at the picturesque view of San Francisco. That it didn’t happen is now simply a fact of life in the Bay, but we need not live with this fact.

Two weeks ago, I tackled the science of traffic congestion: why it happens and the damage it does to our transportation system [1]. Last week, I examined the best way to cope with congestion – an anti-congestion toll – and how to craft anti-congestion policies with an eye toward equity in Marin [2].

Yet an anti-congestion tolling plan works best when there are effective alternatives. Golden Gate Bus is, when it doesn’t face traffic, a fast and efficient mode of transportation. Unfortunately, it doesn’t run often and it runs into traffic frequently, especially within The City. As well, transfers between it and other long-distance transportation systems like BART and Caltrain are poor at best. Marin and Sonoma residents, then, don’t have viable alternatives to driving if their destination isn’t downtown San Francisco.

The plan in a nutshell

SMART South, as I call this plan to differentiate it from the SMART that is currently under construction, consists of the completion of the SMART system, the upgrades needed to operate at eight trains per hour per direction, and the upgrades, construction, and new trains needed to run SMART south to San Francisco and as a subway under Geary Boulevard.

This will include electrification of the existing line ($125 million [3] for 39 miles of track), pushing SMART north to Healdsburg and south to Marin City ($537 million [i] for 40 miles of surface track), rebuilding the Alto Tunnel ($60 million [4]), tunneling from Marin City to the Golden Gate Bridge ($850 million [ii] for 1.7 miles of tunnel), retrofitting the bridge ($392 million [5]), tunneling from there to the Geary Boulevard Subway ($1,365 million [ii] for 2.7 miles; the Geary Subway segment would be built separately from this project) adding passing track to the existing and new surface line to allow for higher frequency ($705 million [i] for 52 miles), elevating the downtown San Rafael track ($93 million [ii] for 0.6 miles of track), finishing the bike path from Marin City to Cloverdale ($5 million [i]), adding the Corte Madera/Larkspur, Mill Valley, Marin City/Sausalito, Vista Point, and Presidio stations ($500 million [ii]), and buying Caltrain-compatible bi-level electric trains to allow for a much and more frequent longer line ($654 million [6] for 38 3-car trains). There are also $250 million in miscellaneous costs associated with finishing the SMART line as promised. In total, this comes to about $5.5 billion for an effectively-new 85-mile system or about $65 million per mile. That’s quite a steal for American construction costs.

Passengers would be able to transfer at Union Square to BART and Muni’s under-construction Central Subway, and at the Transbay Terminal to Caltrain and high-speed rail. Depending on how the second transbay tube is built, passengers would be able to move on either to downtown Oakland or south to the Oakland Airport and Fremont without transferring.


The short of it is that anti-congestion charges on the Richmond-San Rafael Bridge (RSRB), Golden Gate Bridge (GGB) and the Novato Narrows could be leveraged into $1.4 billion [iii]. A 0.4% sales tax in Sonoma, Marin, and San Francisco could be leveraged into another $1.3 billion [iv], for a total of $2.7 billion raised locally. Regional, state, and federal monies cover about half the cost of major transit construction projects in California, and that’s the balance of the project. Ongoing operations & maintenance would be covered by fares, tolls, and sales tax income as well as state and federal support. The gritty details of all this, including a speculative financing plan and how to calculate anti-congestion tolls, will be in an upcoming post for the people who are truly interested.


SMART South as envisioned would be built alongside a secondary transbay project extending Caltrain and some kind of heavy rail subway across the bay to Alameda County. These would happen regardless of SMART South, so its value is really in connecting the North Bay with the South and East. How many riders would use that link?

Last year, I examined what higher frequencies could do for the base SMART system’s ridership and arrived at a conservative 12,100 daily trips. Adding the southern Marin, downtown Novato, and Presidio stations would push the ridership up to 22,000 using the same model [7].

Using 2014 Census data, it appears that roughly 7,500 people live within a half-mile of SMART and SMART South stations and work within a half-mile of BART or Caltrain, or vice-versa [8]. While only half of them are likely to commute via transit after this extension, additional riders would come from outside the half-mile radius, and commutes only account for about 20 percent of all trips. Wrap all that up and we have another 50,000-80,000 trips per day, depending on fares, travel time, and transfers. Given that this range comes from a model that assumes no anti-congestion tolls - and that such tolls boost transit use – I’d lean more towards the higher than the lower number. That would nearly double the capacity of the Golden Gate and northern Highway 101 corridor.


So for about $5.5 billion, paid for with new tolls and a 0.4% sales tax, SMART could dip south into San Francisco, fully tying Sonoma and Marin into the regional and statewide rail systems. It would provide a viable alternative to the freeway. Alongside tolling, it would make congestion a thing of the past in the North Bay, making deliveries more timely and commutes much more reliable.

This is merely a draft, of course, and should be subjected to more rigorous study. The political difficulties of yet another sales tax and yet-higher tolls are apparent. But SMART South is eminently attainable. With a little ingenuity and a little optimism about the future, it could be our next stop.


[i] Estimate from existing SMART costs.

[ii] Estimate from a variety of peer projects.

[iii] Assuming borrowing is on the same terms as MTC’s toll-related bonds and is spaced according to construction needs.

[iv] Assuming borrowing is on the same terms as SMART’s sales-tax backed bonds.

Works Cited

[1] David Edmondson, “The Science of Traffic and Its Awfulness,” The Greater Marin, July 3, 2017.

[2] David Edmondson, “Let’s Get into the Weeds: A Congestion Charge Plan for Marin,” The Greater Marin, July 10, 2017.

[3] Stephen A. Gazillo, “A Planner’s Guide to Fixed Guideway Electrification Projects,” Transportation Planning, November 2005.

[4] County of Marin, “Investigative Study to Begin on Alto Tunnel” (County of Marin, January 10, 2017).

[5] Charles Seim, Mark Ketchum, and T.Y. Lin International, “Golden Gate Bridge Mass Transit Feasibility Study” (San Francisco, CA: Golden Gate Bridge, Highway and Transportation District, October 1990).

[6] Railway Gazette, “Caltrain Signs Double-Deck EMU and Electrification Contracts,” Railway Gazette, August 16, 2016.

[7] Transportation Research Board, “Elements Needed to Create High Ridership Transit Systems,” Transit Cooperative Research Program (Washington, DC: Federal Transit Administration, March 2007).

[8] Center for Economic Studies, “Longitudinal Employer-Household Dynamics” (Washington, DC: United States Census, 2014).

Let’s get into the weeds: a congestion charge plan for Marin

Last week, we discussed how tolling designed to eliminate congestion would improve travel times and improve the efficiency of our roads. [1] This week, we get a bit into the weeds about how such a plan would work.

How to eliminate congestion in the North Bay

Marin has four main entrances, three of which have significant rush-hour congestion: the Novato Narrows, Highway 37, the Richmond-San Rafael Bridge (RSRB), and the Golden Gate Bridge (GGB), with Highway 37 being the odd one out of the congestion game.* These congested gateways are a drain on people’s time and public resources. The Narrows is only moving 85 percent of the vehicles per hour that it could, the RSRB moves just 61 percent, and the GGB moves just 74 percent capacity.**

The entrances are all chokepoints, with few or no parallel routes, so they are ideal for tolling. The toll need to be charged both directions, so existing tolls on the GGB and RSRB would be cut in half: half heading inbound and half outbound.

A toll that eliminates congestion would allow another 4,380 vehicles per hour to move at rush hour, a staggering 35 percent increase over today's numbers.

To determine the exact amount a congestion toll ought to be, we need some pieces of information: the maximum number of vehicles per hour each road can move, also known as their flow rate; the existing flow at peak hour; the amount of demand there is at peak hours; how long the flow is congested; the length of the congestion; and the median income of the area. If all of this is known, then a little calculus is needed and then, voilà, the result. Unfortunately, with publicly available data it is very difficult to determine demand, so we need to work a little bit harder for a less precise number.

I wanted to find out how much extra time people were spending in traffic than they would if all the roads were at their maximum flow and how much that lost time is worth. Tie these elements together and we get some estimates. The Narrows needs a charge of about $5 ($108 extra per month), the RSRB needs $5 on top of its existing toll (, and the GGB needs $6 on top of its existing toll. Combined, these tolls would bring in an additional $102.4 million per year, which could be used to secure bonds of up to $1.3 billion (over 30 years at a 4 percent interest rate). For context, the Golden Gate Bridge brought in $129.5 million in toll revenue in the 2015/2016 fiscal year. [2]

Look regional

Because Marin’s transit system is part of a broader regional system of bridges and highways, any tolling plan would need to be regional in order to ensure that traffic doesn’t spill over onto some other road. As previously mentioned, Marin’s entrances have very few good roads and therefore can be fairly easily tolled. However, if Marin manages to unclog its traffic, East Bay commuters might try to cut through the county on their way to San Francisco. More drivers on Marin’s roads means higher tolls, and then Marinites and Sonomans would have to pay more than before. Any toll plan would need to be implemented across the whole region to ensure all the region’s freeways are used as efficiently as possible and nobody is charged too much.

Build equity into the toll

One of the oft-cited problems with charges like this is that it is regressive as it hurts the poor, who have more time than money, more than the rich, who have more money than time. I outlined a solution briefly last time, [3] but here it is worth fleshing out a bit more.

Carbon taxes are often structured in such a way so as to be revenue neutral, [4] so whatever income is generated is refunded to the users. However, an unclog-the-roads toll, like what’s proposed here, should be used to increase the tolled travel corridor’s capacity and so needs to provide a useful income. This could be done either by selectively exempting vehicles registered to poor drivers, but that would be logistically difficult. Instead, tolls should be a refundable state tax credit that phases out as income increases. (Refundable tax credits are refunded to the taxpayer whether or not she owes any money, and so they are always returned.) FasTrak, or whoever the vendor is, would inform their registered users every year how much they paid in congestion tolls, which the users could report on their taxes. Non-registered users, like tourists or infrequent users, would not get a form so they’d end up paying no matter their income.

It’s difficult to estimate how much of the toll income would be refunded, but given that drivers tend to be wealthier in general, especially in the suburbs, this may not be a significant portion of toll revenue. Should it become viable to issue monthly rather than annual refunds to poor drivers, it would likely encourage driving, which would in turn increase congestion. That would require higher tolls, offsetting the revenue lost to refunds.

Regardless, because the toll essentially is trading time for money, and the amount of time currently used up in traffic is worth about $102 million per year, that is roughly how much would come in from these tolls.

Next time, we will consider just what might be done with this revenue stream to make travelers lives better.


* Traffic from 37 does add to congestion on Highway 101 to the south, and congestion does regularly crop up on the road, but the MTC analysis this post relies on [5] reports very little congestion on the road itself. Without a proper traffic management analysis, it would be tough to say how these tolls would actually shift congestion around the North Bay.

** This is current peak-hour flow divided by optimal traffic flow. Optimal traffic flow is the maximum capacity of highway lanes [6]; peak-hour flow is from Caltrans [7].

Works Cited

[1] David Edmondson, “The Science of Traffic and Its Awfulness,” The Greater Marin, July 3, 2017.

[2] Golden Gate Bridge, Highway, and Transportation District, “Comprehensive Annual Financial Report,” Annual Report (San Francisco, CA: Golden Gate Bridge, Highway and Transportation District, 2016).

[3] Edmondson, “The Science of Traffic and Its Awfulness.”

[4] UNFCCC, “Revenue-Neutral Carbon Tax,” Momentum for Change, accessed July 4, 2017.

[5] Metropolitan Transportation Commission, “Time Spent in Congestion,” Vital Signs, accessed April 29, 2017.

[6] John D. Zegeer et al., “Default Values for Highway Capacity and Level of Service Analyses,” National Cooperative Highway Research Program (Washington, DC: Transportation Research Board, 2008).

[7] Division of Traffic Operations, “2015 Traffic Volumes on California State Highways” (Sacramento, CA: Caltrans, 2015).

The science of traffic and its awfulness

If only the guy in front of you would drive better, perhaps you wouldn’t be stuck in traffic. Slow, fast, break. Slow, fast, break. Shift lanes. Whoever it is, that driver sucks.

That’s what the driver behind you thinks about you, and what the driver behind that driver thinks about him, and so on. It’s traffic congestion, and it’s bad for everyone.

There are two aspects of congestion to be explored here in very brief detail: its physical results and its economic causes. Only once we understand the economic causes will we be able to figure out a solution.

Traffic makes roads terrible

Roads move traffic at a rate governed by a very simple model: the density of the vehicles per mile times their speed. But there is a trade-off. As the density of vehicles goes up, their speed goes down as drivers become more cautious. There’s a balancing point where density and speed allow a road to move the most vehicles. Shift the density a bit either way, and you don’t move as many vehicles, and you’re not getting your money’s worth out of the road. The relationship looks like this:

When traffic gets bad, roads exist on right-hand side of the diagram. Drivers like being on the left-hand side where speed is pretty high,, and taxpayers should want to be right in the middle, getting the most for their money. Image from ACCESS magazine [1]

When traffic gets bad, roads exist on right-hand side of the diagram. Drivers like being on the left-hand side where speed is pretty high,, and taxpayers should want to be right in the middle, getting the most for their money. Image from ACCESS magazine [1]

So, traffic congestion is when the road is operating somewhere on the right side of the curve. Basically, at rush hour, there are too many people that want to use the road all at once, increasing density way past the efficient level, making everyone slow down and the roads awful. This video does a good job of showing how this works. It’s boring, but stick with it [2]:

Basically, everyone slows everyone else down by trying to use the road all at the same time. Every additional vehicle on the road slows everyone else down just a little bit, which adds up to a lot. Past the balancing point, each additional vehicle also makes the road less efficient. Just how much is something we can quantify.

Time is money

When someone is slowed down in traffic, they get to wherever they’re going later than they’d hoped. This might cost maybe a few minutes, but when multiplied out over everyone stuck in (and causing) that traffic it ends up being a lot of person-hours. Economists, measuring a whole host of factors [3], translate this time cost into a monetary cost. Every vehicle “pays” a certain amount of time (alongside gas, maintenance, and tolls) which is the average cost of using the road. When someone new comes into the road, that average cost goes up just a little bit at the margins. This is called marginal cost of using the road. Each driver feels only the average cost; the marginal cost is divided up evenly between everyone else around them on the road. It’s like when someone cuts in line: everyone behind them is slowed down a bit.

In short, every vehicle faces the average cost and forces everyone else to “pay” a little bit more. Brendan O’Flaherty writes:

The final major kind of external cost that drivers impose on each other is congestion. On an otherwise deserted country road, or even on most city streets at four o'clock in the morning, it's difficult for one car to get in the way of another car or to impede its progress, and so congestion is not a problem. But when tens of thousands of cars an hour are converging on the Lincoln Tunnel during rush hour, they get in each others' way, and since the number of cars that can go through the tunnel in any minute is less than the number that want to go through, each car that enters the tunnel during rush hour is forcing all the cars behind it in the queue to wait a few seconds. Because drivers don't bear the cost of the congestion they cause, they cause too much of it.
I'm not saying that drivers don't bear the costs of congestion. Of course they do; being stuck in a traffic jam is unpleasant and time-consuming. But I don't bear the costs of the congestion I cause; instead I bear the costs of the congestion you cause, and so you have no incentive not to cause it. Or more starkly, think about a line of cars going through the Lincoln Tunnel. The first car may encounter no congestion, but if it were vaporized, every other car would get through the tunnel more quickly; and so the first car is causing a lot of congestion. The last car, by contrast, is encountering a lot of congestion, but causing none (if it were vaporized no one else would get through the tunnel sooner). [4]

O’Flaherty goes on to also describe how people shift their schedule around to avoid traffic, which has its own cost, and how people sometimes just grin and bear it, choosing to drive in traffic and pay that cost on the road instead. [5]

But all this cost is just in terms of time. It’s wasted, lost to everyone. If there were a way to shift that cost into dollars, then it would be possible to put that cost to productive use, or to return it to the people who need money more than time.

Anti-congestion toll

This is where congestion pricing, or demand pricing, or surge pricing, or whatever you want to call it, comes in. In essence, it’s a toll that charges people what they would otherwise pay in terms of time as well as the cost they impose on other users of the road (the marginal cost). This smooths out demand, so the road never gets clogged, and allows the money to be put to better use. The tolls collected could also be returned to poor drivers as a refundable tax credit. (I would have the electronic toll collector contractor, i.e. FasTrak, mail out a tax form to each of its customers itemizing the amount paid in anti-congestion tolls by month, which the customer could report on their taxes as a credit contingent upon income.)

The goal would not be to return the road to free-flow speed, but rather to the balancing point, where a road moves the most cars per hour possible. More people would therefore be able to get through the system at rush hour, meaning the whole driving system would be more efficient. Any money not returned as a tax credit could sponsor public transit or road improvements.

The London tolling plan, which did not include the tax credit, saw travel times fall and then stabilize within the controlled area, though traffic volumes continued to fall significantly. [6] London also saw the number of crashes and traffic deaths fall significantly. [7] The Stockholm plan saw the similar results but, intriguingly, drivers who were interviewed after the toll went into effect didn’t think they had changed their travel behavior. [8]

This is the only way to ensure congestion doesn’t occur without a recession because, as discussed above, people normally don’t feel the cost they impose on everyone else on the road and so they overuse it. With a toll like this, people will feel the full cost of their road use and so allow society to get the most out of the investment in roads.

Of course, people don’t like to be charged for what they previously got for free, especially when that would represent an unknown change. Stockholm residents, for instance, hated the idea of a congestion charge until it was attempted. [9] Manhattan has tried for years to implement a charge only to be blocked by state lawmakers. [10] When San Francisco talked about doing something on the Golden Gate Bridge, then-San Rafael mayor Al Boro called it a “Marin commuter tax.” [11]

Nevertheless, given that it is the only way to permanently resolve traffic, it is worth exploring how big the charge would be and what is possible with that income. That’s what we will explore next time.

Works Cited

[1] Richard Arnott, “A Bathtub Model of Downtown Traffic Congestion,” ACCESS Magazine, June 1, 2015.

[2] Washington State Department of Transportation, Doug MacDonald - Rice and Traffic Congestion, 2007.

[3] J. Hewitt, “The Calculation of Congestion Taxes on Roads,” Economica 31, no. 121 (1964): 72–81, doi:10.2307/2550927; Roberto Ayala, “The Value of Travel Time Savings: Departmental Guidance for Conducting Economic Evaluations Revision 2 (2014 Update)” (Washington, DC: US Department of Transportation, July 9, 2014).

[4] Brendan O’Flaherty, “Congestion,” in City Economics, 1st ed (Cambridge, Mass: Harvard University Press, 2005), 52.

[5] Ibid., 56.

[6] Transport for London, “Public and Stakeholder Consultation on a Variation Order to Modify the Congestion Charging Scheme Impact Assessment” (London, UK, January 2014).

[7] Alex Davies, “London’s Congestion Pricing Plan Is Saving Lives,” Website Type, Wired, (March 10, 2015).

[8] Jonas Eliasson, How to Solve Traffic Jams (Lausanne, Switzerland: TEDxHelvetia, 2012).

[9] Ibid.

[10] Ben Fried, “Factchecking Cuomo’s Revisionist History of NYC Road Pricing,” Streetsblog New York City, February 18, 2015.

[11] Michael Cabanatuan, “Bridge Raises Tolls, Denies Doyle Dr. Funds,” SFGate, March 15, 2008.

Polishing poop: the SMART schedule

Update: SMART just reached out to me with news that they are looking at the schedule again based upon the feedback they've received, so stay tuned for a new schedule.

By now, word of the final SMART schedule is out, and the response is relatively negative. The schedule (PDF here) has a 90-minute gap right at the heart of the rush hour, between 7:26am and 8:56am. Folks were not having it on Twitter. These two tweets are fairly representative:

The SMART Board wasn’t happy with the results either. San Rafael Mayor Gary Phillips told staff at the meeting where the schedule was released, “With these kind of gaps I’m concerned people will say, ‘I’ll continue to drive my car,’ I would encourage we revisit this” (1).

For their part, SMART spokeswoman Jeanne Belding wrote me indicating the schedule was a result of a survey of potential riders after a draft schedule was released. She added, “The schedule we are opening with is based on previous feedback, our staffing, and takes into account the fact that we have a single-track system” (2).

SMART General Manager Farhad Mansourian indicated that the reason for the schedule is, in short, that it is complicated: with the system running as it is, a change to one train means all trains must be changed. And there’s freight trains to deal with, only four active passenger trains, and so on (1). But I think there is more to discover in the schedule, and that the system is not nearly as hamstrung as GM Mansourian seems to indicate.

Polishing poop

Let it be known that just because the schedule is bad doesn't mean SMART isn't operating under real constraints. Undoubtedly, this was someone's attempt to polish the poop (or shine the shit) that is their situation. Nevertheless, this is definitely not sufficiently shiny yet, and I think we can do better. Let's look under the hood at what we're dealing with.

First, we can divide the schedule into a series of what I call platoons: four trains, each departing 30 minutes after one another, followed by a minimum of a 90-minute break, followed by another platoon. These are set by the number of trains SMART can run at any given time: just four of their seven-train fleet. Spokeswoman Belding told me they have had trouble hiring drivers, citing the high cost of living as the reason, and so they are limited to just four trains on the tracks at any given time. She did not answer a question about when they expected to be at full strength (2).

Another limiting factor are the passing tracks. Taking best practices from Switzerland and Germany, SMART built a single-track commuter system with four four-mile-long passing tracks at key locations so trains could run in both directions and pass one another. The problem, of course, is that these passing tracks limit the schedule to certain ranges. What those ranges are, SMART can’t say, but we can assume that the schedule as presented is one solution.

Taking all this into account, we can trust that changing one train’s arrival or departure time more than a few minutes requires the rest of the schedule to be changed by the same amount.


So now, the problem at hand: what to do about that 90-minute morning gap.

I took this as gospel and rotated the entire schedule around the arrival time of Train 5 at downtown San Rafael. Here’s what I call San Rafael Rush:

The San Rafael Rush schedule gets everyone to downtown San Rafael before 9am and ensures people still have time to walk to work after the last train arrives. Unfortunately, it makes things tough on northbound commuters, giving very little flexibility for Marinites working in Sonoma.

The second option is what I call Early Bird:

This option pushes everything back by about 10 minutes, ensuring that the 90-minute gap ends right as rush hour is at its worst. I’m not so keen on Early Bird, as few people arrive at work during those early morning hours served by the first platoon, but it does make sure that people can at least get to the office by 9am.

There are other methods of making the schedule fit, but it’s important to understand that SMART is operating in a tension between what riders say they want versus what the data says they should do. Census data shows what we know: that a plurality of people working in San Rafael arrive between 8am and 9am, people whom the current schedule poorly serves. SMART also insists that its schedule is based upon feedback from its own scheduling poll of potential riders.

It’s also important to note that GM Mansourian’s statements that nothing can be done are obfuscation. True, SMART is limited in its schedule movement, and they will be until they get the drivers they need, but they can indeed do better than what they’ve given. There is a bit of hope on that front. The GM indicated they’d be open to changing the schedule based on feedback after the train has started running. “This is what we will test and get feedback. If we can tweak, we will tweak” (1).

PS - And just because you've heard it a thousand times already, I did promise to remind you not to risk getting hit by a train. Sure, they'll be far less dangerous than drivers, but that doesn't mean you want to be the guy making everyone late for work because you got squashed by a train. Also, I will take this opportunity to repost one of my favorite ads, Dumb Ways to Die.

Works cited

(1) Prado M. SMART schedule has major service gaps. Marin Independent Journal [Internet]. 2017 May 17 [cited 2017 May 22]; Available from: http://www.marinij.com/article/NO/20170517/NEWS/170519805

(2) Belding J. Service schedule & trains. 2017.

Why good transit map design matters

Public transit is not just for people in big cities. All across the United States, public transit is a lifeline for millions of people who cannot afford a car or cannot drive. Unfortunately, these routes often provide the barest minimum of service: four round trips per day, no Sunday service, early end times. To add insult to injury, they are very rarely mapped in a comprehensible way, forcing riders to study dense timetables and obtuse maps.

To solve this problem in Tompkins County, New York, I created a frequency map indicating all stops, all routes, and major connection points.

One page from the rider guide

The existing maps are quite poor: there is no system-wide map, forcing riders to flip between pages to follow a single line along its course. Service levels aren't indicated, with weekend or late-night service shown as being just as important as weekday or high-frequency service. Often, new riders will avoid buses they don't know, even if the bus is going where they want to go, simply because there's no way to know where it will take them. Further, the local system, TCAT, has flag-stop service, where riders can get on or off a bus wherever the driver can safely pull off the road, but it is not indicated. In talking with riders, my team and I found people often didn't even know about the service. The existing maps' final sin is that it groups routes together on the same line even when those routes have nothing in common.

The same area as above, but in my redesign

The new map fixes each of these problems. It adjusts line thickness to indicate mid-day frequency; groups late-night and weekend routes together with their weekday counterpart routes (the 70 is the weekend counterpart of the 30, so these are combined into a single line); indicates flag-stop zones; and is envisioned as showing the full county's system in a single panel. For rural riders, for whom roads are critically important to wayfinding, the road grid is underlaid behind the entirety of the map.

I started a similar project a few years ago for Marin and Sonoma, but my design skills were simply too poor to finish. It taught me a lot, but, given the major changes since starting this project, I suspect I would need to start again. More interesting would be a better map of Bay Area transit service: 

Bad design is not necessarily the fault of the transit agency - good designers are expensive and need management, and often staff and budgets are already too small. There is little time or energy to spearhead change. Still, they ought to consider the benefits that come from comprehensibly conveying the entirety of the system in one clear image.

Election Maps for November, 2016

As anyone who follows American federal politics knows, location can mean a lot when it comes to who votes for whom. Electoral maps on the local level are rather more rare, but no less informative. Here below are the maps of relative support for selected races and candidates from the November 2016 election. (1)

(You can click on the images for a bigger size and the PDFs are available, too, by request. Feel free to request more maps in the comments section. Note that this is not the final tally but one of the interim tallies. The exact levels of support may change a bit.)


Hillary Clinton won every single voting precinct in the county with at least 52 percent of the vote and Donald Trump got second place in all but two precincts, each of which are worth less than 100 voters. Rather than map first and second place, then, I wondered how the voting was distributed across the county.

Tiburon, Blevedere, Ross, Corte Madera and Novato were the most conservative of Marin's cities, though in Marin that means voting for Clinton with 55-65 percent of the vote rather than 80 percent. Jill Stein did best in West Marin's villages and downtown San Rafael, and Trump did his best in Novato.

District 4 Supervisor

In Marin's marquee election, Dennis Rodoni beat out Dominic Grossi in second-round voting. Broadly speaking, Rodoni carried the northeastern half of the district with broad-based support in Woodacre, Dillon Beach, Novato, and a slice of Larkspur. Grossi took the district's southwestern half, getting support from San Rafael, Corte Madera, Mill Valley, and a handful of West Marin villages.

This mirrors the support received in the first round, indicating to me that there was not much convincing going on in the latter half of the campaign.

Proposition 64: Marijuana Legalization

As with Clinton support, no precinct in the county rejected the measure. That said, we can see that support for the measure largely followed the same lines as the presidential race: places that were cooler on Clinton were cooler on legalizing pot, and vice versa.


Measure A: "Strong Start" quarter-cent sales tax for child services

Marin's marquee ballot initiative was much more divisive. The measure, which would have passed a quarter-cent sales tax "to fund expanded preschool, child care and health services for low-income children in Marin," (2) did not pass. Though no precinct had less than half its voters supporting the initiative, because it was a new tax it needed a two-thirds majority. There is some of the liberal/conservative split seen in these results, but at first glance it looks bit more ambiguous than in the presidential election.

Look under the hood, however, and one finds a significant correlation between Clinton support and Measure A support. Here's that scatter-plot:

Basically, for every percentage of Clinton support in a precinct, that precinct also saw a half-point more support for Measure A. That's not a one-to-one match, but it is definitely there.

So Marin does seem to have something of an ideological split. Indeed, it may be possible to use Republican and ballot measure voting patterns to mark which areas are more or less liberal than the Marin County average. What to label this divide, of course, would be quite heated: both sides want to seem mainstream (and both are!) and being "conservative" in Marin is almost seen as a character flaw. Chime in in the comments if you have ideas ("populist" vs "progressive" or something?)

The core message of this post is that Marin does have politically diverse geography beneath the liberal veneer. Understanding it should be integral to any political outreach program.

1. Marin County Elections Department. November 8, 2016 General Election Statement of Votes Cast [Internet]. 2016 [cited 2017 Feb 7]. Available from: http://www.marincounty.org/depts/rv/election-info/past-elections/page-data/tabs-collection/2016/nov-8/sovc-listing-page

2. Halstead R. Marin voters face 11 tax measures on November ballot. Marin Independent Journal [Internet]. 2016 Aug 12 [cited 2017 Feb 7]; Available from: http://www.marinij.com/article/NO/20160812/NEWS/160819921


Returning to the Flood

Last week, San Anselmo came within inches of yet another disastrous flood. Perhaps it’s time to revisit the flood detention basin issue.


After the flood in 2005, a plan was put forward to build a flood detention basin in Memorial Park which would protect the downtown from floods by temporarily diverting water to the park during heavy storms. The town received a $8.72 million state grant to build the $17.2 million basin and planning work began. In 2015, voters rejected Measure E – without downtown’s support – which would have allowed the use of Memorial Park as a basin. The flood zone board is now trying to figure out what to do instead.

Support for Measure E. Image by the Author.

Of those currently on the council, Matt Brown and Ford Greene oppose the conversion of the park. Councilmember Brown led the charge to pass Measure E. In the same election he also ousted now-former council member and basin supporter Doug Kelly. Councilmember Greene has consistently sided with Marin’s populists on issues such as affordable housing, zoning reform, transit, and the flood basin. Both have appeared at Citizen Marin and associated group meetings. The rest of the council (Kay Coleman, Tom McInerney, and John Wright) supported the conversion of the park [1] and presumably still would if it were a possibility.

Councilmember Brown is certainly unmoved by last week’s close call. To quote the IJ, “Brown said he and many other San Anselmo residents wisely chose not to buy property in the floodplain. He said spending millions of dollars to address flooding in the Ross Valley would be ‘tantamount to a bailout of somebody’s real estate decision.’” He also voiced skepticism that the town should even pay for flood mitigation: “[H]ow how much are we willing to pay to put that plan into place? This is a multimillion-dollar project, one that I believe the people have to decide if it’s worth implementing.” [2]

Councilmember Greene doubts that the basin system would even make a difference but advocates instead for another basin in Fairfax and the demolition of 636 San Anselmo Avenue, which straddles the creek. [3] This is one of the proposals currently undergoing public comment.

What now?

The basin would have been part of the work plan for Flood Control District 9, the board charged with reducing flood risk in Ross Valley, but the rest of the work plan is still proceeding. This includes raising bridges in San Anselmo, Ross, and Fairfax; creating flood detention basins at Loma Alta and Lefty Gomez Field in Fairfax; creating additional storage capacity at Phoenix Lake; and improving the creek channel in Ross and Larkspur. [4] The conversion of Lefty Gomez Field has already inspired opposition. [5]

Of course, the Memorial Park basin was part of the overall flood mitigation strategy for the watershed and, with its demise, an alternative needs to be found. There are four alternatives currently before the district board, each of which costs approximately the same as the Memorial Park conversion. You can read all about them here, but the gist of them is that each would be disruptive somewhere. The No Basin Alternative involves heavy roadwork on the Sir Francis Drake Boulevard bridge. The Sleepy Hollow alternative involves a detention basin at Brookside Elementary. Councilmember Greene’s preferred alternative would involve at least partially demolishing two of downtown San Anselmo’s iconic buildings. [6]

It is a shame that downtown San Anselmo has to endure not just neglect from its neighborhoods but active opposition to its success. A town ought to be more than a commodity to be used but a community that sticks together and looks out for one another. Downtown is supposed to be the heart of that community.

Works Cited

[1] Richard Halstead, “Dueling Ballot Measures over Flood Control Divide San Anselmo,” Marin Independent Journal, September 26, 2015; Peter Seidman, “Shelter from the Storm,” Pacific Sun, May 4, 2015.

[2] Richard Halstead, “Ross Valley Flood Scare Brings Calls for Action,” Marin Independent Journal, January 13, 2017.

[3] Ibid.

[4] Marin County Department of Public Works, “FZ9 Project Fact Sheets,” Marin County Watershed Program, accessed January 18, 2017.

[5] “Save Lefty Gomez Field,” accessed January 18, 2017.

[6] Liz Lewis, “Update-Analysis of Replacement Project Alternatives for DWR Grant” (San Anselmo, CA: Flood Control Zone 9 Advisory Board, June 21, 2016).

A high-frequency SMART corridor, revisited

In this week’s IJ Forum, guest and transit skeptic Richard Hall brought up something I didn’t expect: the argument that transit frequency equals freedom [1]. This is something that transit advocates have been harping on in Marin and elsewhere for years, and one of the key factors that is likely to hold back SMART from reaching its potential. Though I addressed this point back in 2012, with the train opening up this year it is worth revisiting how SMART might be able to reach higher frequencies, and how much it might cost.

There is blood in the single-track stone

Once operational, SMART’s trains will run in both directions simultaneously despite having only one track. It can accomplish this through the judicious use of passing tracks, meaning a strategic stretch of the corridor will be double-tracked so trains can pass one another.

At full build-out, from Cloverdale to Larkspur Landing, there will be 4 such passing tracks, each of which is 12 miles long, which will allow trains to run every 30 minutes in either direction. (This 30-minute time is called the “headway” in transit-speak.) To double the frequency to 15 minutes, we would need to double the number of passing tracks to 8. To double frequency again to 7.5 minutes, we would need to double the number of passing tracks again to 16.

At this point, the permanently single-tracked Puerto Suello Tunnel between North and Central San Rafael becomes the choke point. Any further increases would need to widen the Puerto Suello Tunnel and basically double-track the whole system. This would likely overkill for the foreseeable future.

Using the existing construction costs as a guide, the cost for the passing tracks on a 15-minute headway system is $60 million. The cost for a 7.5-minute headway system is twice that: $120 million [2]. If we include the cost of the trains needed to run such frequent service, the total cost would be $120 million and $307 million, respectively [3].

We can squeeze blood out of this stone if we are willing to spend the money for it.

Of course, ridership at this point becomes a problem. Is it worth it to run trains this often? Definitely.

There are almost 43,000 jobs and over 19,000 people within a half-mile of SMART’s stations. 2,237 people live and work within the station areas [4], but commute trips account for only about 20 percent of all trips, and this doesn’t include people who might use the park and ride services. Ballpark figures, using a ridership model [PDF] from the Transportation Research Board put the ceiling of SMART’s potential ridership at about 25,000 [5]. SMART itself estimates its ridership will be around 4,500 trips per day, which the model agrees with.

We need to dive a little bit into some wonky economics for a moment using these numbers but bear with me.

When a transit operator alters some aspect of a transit service, they of course also alter the ridership. Increase the vehicle speed and ridership goes up. Increase fares and ridership goes down. This is called the elasticity of demand: how much does a given change affect the demand for that transit service. The first example mentioned references the in-vehicle time elasticity because it indicates how much people change their habits based on how much time they spend in a vehicle. The second example is the price elasticity because it deals with the price of using the service.

Headways alter the access time elasticity, which is more valuable to people when it comes to transit. As much as you may enjoy your walk to the train, you probably enjoy the fact that you can work or sleep on the train en route a little bit more.

Access time elasticity among commuters is 1.28, which is pretty high. Decreasing average wait times by 1 percent by decreasing headways by 2 percent increases ridership by 128 percent [6]. Doubling SMART’s frequency to 15 minutes, then, will get us a 64 percent* increase in passengers: from 4,500 trips per day to about 7,380. Doubling frequency again could get us to 12,100 trips per day. This, of course, does not count the number of people who would use SMART for non-commute trips too, and it also ignores the effects of improving on the very infrequent service SMART is planning on providing in the middle of the day.

I would argue that a 15-minute frequency would absolutely be worthwhile on the basis of the commuters alone (the cost per new commute trip is $42,000, half the cost per trip on the baseline SMART system). A 7.5-minute system is not as cost-effective based on new commuters, but is significant if we include non-commute trips.

It may also be a viable alternative to a wider Novato Narrows. The traffic congestion there is in part due to a 15 percent increase in the number of cars travelling through to Central San Rafael, or roughly 500 more vehicles per hour [7]. Diverting 3,690 trips per rush hour (half of 7,380) would alleviate that congestion, at least until drivers fill up the space again thanks to induced demand [8].

So Hall is quite right on this point: transit frequency influences how people travel and how many people use the system. I am unsure whether he knew just how influential frequency can be, but no matter. SMART would do well to examine the effects of increasing its service frequency.

* Average wait times are half of the headway: if you arrive at a random point during the inter-train period, your average wait is half the full headway. Doubling frequency, then – a 100 percent increase – decreases average wait times by half that, or 50 percent.

Note: Given how off-topic any comments section can get in a SMART-focused article, I have turned them off for this post.

Works Cited

[1] Traffic in Marin, IJ Forums (San Rafael, CA, 2016).

[2] David Edmondson, “High SMART Frequency on the Cheap,” The Greater Marin, August 8, 2012.

[3] David Edmondson, “Can SMART Double-Track?,” The Greater Marin, August 6, 2012.

[4] Center for Economic Studies, “Longitudinal Employer-Household Dynamics” (United States Census, n.d.).

[5] Daniel G. Chatman et al., “Making Effective Fixed Guideway Transit Investments: Indicators of Success” (Washington, DC: Transportation Research Board, 2014).

[6] Arthur O’Sullivan, Urban Economics, 8th ed (New York, NY: McGraw-Hill/Irwin, 2012), 295.

[7] David Edmondson, “The 101 Corridor: Transportation Myopia in Practice,” The Greater Marin, January 13, 2013; Caltrans, “Traffic Volumes on California State Highways” (Sacramento, CA: Government of California, 2014).

[8] Connor Jones, “The Street Economics of Induced Demand,” The Greater Marin, December 21, 2015.

Supervisor races were all about geography

The June 7 primary election gave us political junkies a much-needed look at temperature of Marin County's electorate. In lieu of polls, it's easy to imagine the loudest voices are also the most politically powerful, but that doesn't seem to be the case this year. Instead, geography was destiny, with incumbents swinging to large victories in incorporated areas and challengers finding difficulty breaking out of their unincorporated enclaves. In West Marin, it was a classic case of North vs. South.

These maps were created using the First Count data released by the elections office on June 10. Full data counts won't be available until next week. A brief update will outline what changed once these data are released.

District 2: Katie Rice, Kevin Haroff, Frank Egger

Image by the author. Click to enlarge.

This election pitted incumbent Katie Rice against Larkspur councilmember Kevin Haroff and former Fairfax councilmember (now Ross Valley Sanitary District board member) Frank Egger. Haroff was endorsed by the Citizen Marin coalition and came out early against the Larkspur Station Area Plan. Egger orchestrated rezoning large chunks of downtown Fairfax to include more parking a few years ago and has called for more parking around Larkspur Landing.

First, Rice won every precinct and won the district with a whopping 57 percent. Nowhere did she earn fewer votes than her challengers. However, these two split the Citizen Marin vote, and there are areas where Rice won less than the combined totals of her challengers.

Rice was strongest in Gerstle Park, downtown San Rafael, and Larkspur, but would have likely lost Ross, Greenbrae, and northern Fairfax had she only faced one challenger. Egger was strongest in Fairfax, as he often is, but failed to make a good showing elsewhere. Haroff showed strongest in Ross and Greenbrae but did not do so well in his home city.

District 3: Kate Sears and Susan Kirsch

Image by the author. Click to enlarge.

Incumbent Kate Sears beat back a vigorous challenge from Citizen Marin co-founder Susan Kirsch who arguably started her campaign almost two years ago.

Strawberry vigorously pushed back against a then-years-old Priority Development Area, arguing it was a way for Sears to ram through new housing. Though that anger and resentment are largely in the past, its marks remain on this electoral stage, with Kirsch handily winning the Strawberry Peninsula and narrowly earning the support of nearby unincorporated neighborhoods on the Tiburon Peninsula.

Tam Valley/Homestead Valley/Almonte, home to Citizen Marin ally Sustainable TamAlmonte and numerous protests over Muir Woods tourists, however, went to Sears, as did all the incorporated towns in the district. Sears’ strongest support came from Marin City, possibly indicating that attempts to make inroads there by Kirsch and Community Venture Partners’ Bob Silvestri have not been terribly successful electorally. Sausalito, home to Citizen Marin allies and councilwomen Linda Pfeifer and Jill Hoffman, also went strongly for Sears.

District 4: The West Marin melee

Image by the author.

The chaotic 8-way race to replace Steven Kinsey resulted in a runoff between first-place Dennis Rodoni and second-place Dominic Grossi. No doubt the various endorsing bodies will be looking hard at the positions of both as progressive Wendi Kallins (frequently misspelled Wendy) and Citizen Marinite Al Dugan both failed to make the top two.

Rodoni, Grossi, and Kallins, along with 8th-place finisher Mari Tamburo, each claimed victory in at least one precinct. Rodoni’s support was concentrated in the urban areas of District 4, such as Larkspur and Novato; Grossi won the rural ranchlands of West Marin, as is befitting a rancher; and Kallins won her home of Forest Knolls, Olema, and San Geronimo. Dugan’s support, such as it was, came from East Marin, implying his platform of traffic and pension reform just don’t resonate out West.

EDIT: [A typo in my map reversed Rodoni and Grossi, which gives us an interesting chance at counterfactuals. Here's my alternative, corrected analysis:]

Rodoni, Grossi, and Kallins, along with 8th-place finisher Mari Tamburo, each claimed victory in at elast one precinct. Rodoni's support was concentrated in the township and urban areas of southern West and East Marin, including his home of Olema. Grossi won his home city of Novato and the more agriculturally-focused areas of northern West Marin, with just a few pockets of support in Southern Marin. In this first-round of results - we won't have the final vote tally until next week - Kallins won the central areas of West Marin in the San Geronimo Valley. Tam Valley was fairly evenly split.

Dugan support was strongly focused in East Marin, though he came in a distant fourth place. His support, such as it was, indicates his platform of traffic and pension reform just didn't resonate out West.

Intriguingly, Kallins seemed to be most in competition with Grossi, not Dugan, with a negative correlation between Grossi and Kallins vote shares. Dugan seems to have been a candidate on his own, with no clear negative correlation between his vote share and anyone else’s. He shares this in common with the other also-rans and Rodoni. This likely means that Grossi and Kallins split support. If Kallins supporters swing over to Grossi, he would be a formidible foe, able to command support of most of the incorporated areas of his district as well as the more populous pieces of West Marin. Rodoni would need to pick up not just Dugan supporters but many of the other also-rans to match. We will find out more as the Grossi and Rodoni campaigns gear up for the runoff in November.

If a pattern can be drawn, it is that Marinites reward competence and presence in their supervisors. District 4 residents rewarded supervisors with West Marin sensibilities. In Districts 2 and 3, their efforts or not, Kate Sears and Katie Rice have both been at the forefront of efforts to fix their districts' traffic. Yet this also cuts the other way: Strawberry and Greenbrae feel sidelined by their respective supervisors. Building trust will be difficult in these communities, but will be important: the superior organizing power of anger and aggrievement can make governing difficult even for an electorally safe politician.

Author's note: I am digitizing a huge amount of electoral data going back to 2013 for all races. If there is a race you would like to see mapped, let me know in the comments.

Author's other note: Unfortunately, Marin’s elections office does not keep shapefiles of electoral precincts, only lists of which residential properties belong to which precinct. This makes for unpleasant-looking maps, with holes and gaps where roads or uninhabited parcels are. But, short of redrawing hundreds of electoral precincts, it’s the best we have.

Citizen Marin slate loses big – what does it mean for their coalition?

In the aftermath of elections in three supervisorial races where incumbents prevailed handsomely and West Marin’s Al Dugan notched a dismal fourth place, behind progressive third-place finisher Wendi Kallins, it is worth asking whether the coalition that birthed them still has steam.

Citizen Marin and its coalition – Community Venture Partners, MAD, Larkspur Strikes Back, Citizens for Sustainable Pension Plans, and Sustainable TamAlmonte, among others – came to Marin in a splash back in 2011 and has notched up significant victories halting plans that would have allowed Marin’s affordable housing supply to expand in Strawberry, Tam Valley, Larkspur, Terra Linda, and Marinwood. They elected Damon Connolly to the Board of Supervisors and fought hard against new bicycle lanes, transit infrastructure, and zoning reforms throughout the county.

This year seemed like a golden chance for them to solidify a majority on the Board of Supervisors, allowing them to not just block reforms and homes but also entrench car-oriented policies and push against regional planning efforts from a position of power. In Kevin Haroff, they had a Larkspur councilmember; in Susan Kirsch, a longtime activist with name recognition; in Al Dugan, a resume sure to appeal to pension reformers.

They also had time. Talk of the supervisorial races was active in anti-housing circles as early as two years ago, with Susan Kirsch penning a coming-out Marin Voice in 2014.[1] They even felt confident enough to campaign against Measure AA, with Kirsch calling it taxation without representation.[2]

Yet each, sticking to the familiar talking points about housing, regionalism, and traffic, lost big, with incumbents Katie Rice and Kate Sears beating challengers by double-digit margins and Dugan not even notching above 10 percent.

In the absence of solid data, it’s tough to say exactly why the three failed, but it’s likely the anger that propelled the anti-housing coalition to prominence is subsiding in the population at large. Dick Spotswood, for what it’s worth, agrees, writing, “Much of America is angry. Not Marin.”[3]

Even if this is the case, housing advocates and other progressive reformers have only started to win a conversation set by Citizen Marin. They have yet to really start a new one. Supporting the progressive vision of Marin as a collection of welcoming, car-optional and quiet suburban towns is quite a different thing than just disbelieving Citizen Marin’s fearfulness.

And there are real problems that have languished as progressives have fought a drag-out fight for better affordable housing, especially the utter mismanagement of Golden Gate Transit.

There is no guarantee that the conversation will shift or that the coalition is on its way out. It still has elected supporters throughout Marin, and the upcoming Plan Bay Area meetings offer it an opportunity to reignite the anger that launched its march. Dick Spotswood still occupies a powerful soapbox. Yet the tide that nearly shifted Marin’s political landscape certainly feels as though it is ebbing. It simply remains to be seen whether the coalition will be able to launch another successful anti-incumbent challenge.

Works Cited

[1] Susan Kirsch, “Building on Voters’ Push for Political Affirmation,Marin Independent Journal, November 12, 2014.

[2] Susan Kirsch, “Bay Tax Is Taxation without Representation for Marin,” Marin Independent Journal, March 9, 2016.

[3] Dick Spotswood, “Marin Results Don’t Reflect Much Voter Anger,Marin Independent Journal, June 11, 2016.

Bad shuttle routing will make SMART's last-mile problem worse

Shuttle service to SMART’s temporary north end will be winding, slow, and inefficient. That’s bad news for Sonoma’s towns, which are already under strain from car commuting.

Waiting for the Bus. Image by Franck Michel, on Flickr.

Waiting for the Bus. Image by Franck Michel, on Flickr.

Sonoma County is getting passenger rail service for the first time in decades, hopefully starting at the tail end of 2016, with the opening of the SMART train.

SMART is starting to move from being an agency building a train, to an agency that will *run* a train. Big difference. Based on a lot of things I have heard in the past, there is concern that General Manager Farhad Mansourian is a better project manager CEO than an operations CEO. Time will tell, but as we approach the opening and decisions start to fall in place, I am going to reload this blog and follow the topic.

At last month's SMART board meeting, there was a presentation on First/Last Mile connectivity. The PDF is linked above, there is also avideo of this meeting - March 2, 2016 which is interesting and a bit illuminating. The board rightly gives the staff of SMART and the related agencies credit for a lot of hard work, but there are a lot of holes in the strategy which underscores that transit agencies and boards don't really focus test their ideas. They look at a problem, place themselves in the problem, and imagine how the problem needs to be solved. Witness Caltrain discussing workers who can go in "later" because they have "flexible" schedules. They go to work 8-5, my office doesn't even turn the lights on until 9 AM and people get upset at meetings before 10. The schedule isn't "flexible", it's flat out different.

With SMART, the board (mostly local politicians) and staff are working from a very "how do I get to San Francisco" mindset. I can't blame them, a lot of the public reacts this way as well, if you read internet comment boards. Officially, SMART diverges from this message on their website:

Today, more than 75% of commuters in the North Bay travel either within or between the two counties to get to work.

Thus we get to my first topic on the presentation on first/last mile - the North County Coordination to be provided by Sonoma County Transit. This connection is at the direction of the board and not negotiable - Windsor, Healdsburg, and Cloverdale were on the original proposed train line, and have had train service delayed indefinitely due to lack of funding. There is a bus from Cloverdale to Santa Rosa - route 60 - but it's slow compared to the freeway and doesn't go to the train depot on Airport Rd. SCT is going to add a shuttle - as seen in the first/last mile presentation, that will express between the 3 towns and the North SMART terminus at Airport Road. It will mostly run on US-101 in uncongested areas making it a quick connector. Sort of.

The proposed schedule shows the "go south in the AM, north in the PM" mindset of SMART. The only shuttles in the AM run North to South to meet trains, there are no proposed shuttles the opposite direction. One problem - the City of Healdsburg is a net *importer* of labor, not an exporter! SMART is trying to figure out how to get a small population of Healdsburg residents south, instead of the large population that is trying to get TO Healdsburg. The population shift during the day is such that tiny Healdsburg is starting to have parking wars and is now considering using valuable downtown land to build more parking.

Healdsburg has over 4,500 in-commuters and over 3,700 out-commuters. This relatively large shift in population for a town of just 11,000 during the workday is exacerbating housing and parking constraints.

Healdsburg has over 4,500 in-commuters and over 3,700 out-commuters. This relatively large shift in population for a town of just 11,000 during the workday is exacerbating housing and parking constraints.

Housing prices in Healdsburg are forcing the town's workforce to leave the city for Santa Rosa and Cloverdale, less expensive areas, which is bad on its own, but exacerbates the parking problem which leads to bad land use decisions which feedback to make the housing problem worse. That workforce could theoretically take SMART to the Airport and hop a shuttle to town, but it won't exist.

Not only does a northbound shuttle not exist, but the market of workers in Cloverdale who could take the shuttle to Healdsburg will be poorly served. The express shuttle will go from Cloverdale to Healdsburg in 20 minutes. SCT route 60 takes ~40 minutes to make the same run - the express cuts the trip in half and could attract new riders. However, because SMART and SCT are only thinking about "get people to the train" - the stop is located at the decrepit Healdsburg train depot on the outskirts of town (and they are building a $1 Million parking lot there), producing a walk for people making that trip which eats up any time savings. It also means that any tourists from SF who decide to take this route get dumped off in the middle of nowhere instead of the middle of town. Might as well drive. This in order to provide park and ride service to a bus for Healdsburg residents? The Healdsburg depot is out of the way for most Healdsburg residents, the bus will make an additional out of the way stop at the Windsor "Train Depot", before winding to Airport Road. Summary - any sane Healdsburger with a car will simply drive to the Airport Road Station.

The detour through HBG to get to the old depot, and the similar winding trip in Windsor kills the trip time, reducing any incentive for people coming from Cloverdale to use the shuttle. The buses should make quick stops just off the freeway but close to the downtowns - the Amtrak bus stop at Mill/HBG Ave in Healdsburg, and right off the freeway in Windsor next to where there is a McDonalds. Shuttle service like this relies on speed. Optimizing it is the only chance to get the ridership needed to keep the shuttle going and hopefully support the train. This includes understanding that some of the riders will eschew the train altogether, using the shuttle as a fast intra-North County bus service. And they should provide service in both directions at both peaks.

Original Post: Murphy, John. 2016. “SMART Train - Last Mile Connectivity - North County.” Holier than You Blog. March 31.

Other Works Cited

Healdsburg, CA. 2016. “Cerri Site (Purity Building) Redevelopment Project.Healdsburg, California Official Site. Accessed May 29.

Mason, Clark. 2015. “Mass Evictions in Healdsburg Prompt Cries of Racism.The Press Democrat, July 8.

Michel, Franck. 2014. Waiting for the Bus. Photograph.

Sonoma-Marin Area Rail Transit. 2016a. Board of Directors Meeting: March 4, 2016. Petaluma, CA.

———. 2016b. “First/Last Mile Connection.” Petaluma, CA, March 2.

U.S. Census Bureau. 2016. LODES Data. Longitudinal-Employer Household Dynamics Program.

Anti-urban groups fight to keep Drake congested

MAD: Fighting to make its logo a reality. Image from MAD. [1]

Despite years of arguing new housing will exacerbate traffic congestion and blaming the county’s urbanists for indifference to the problem of congestion, Marin Against Density (MAD) is now fighting against new traffic lanes on Sir Francis Drake (Drake), arguing the project will actually reduce capacity. Fact check: it won’t.

The project

Drake Boulevard is a mess south of Ross from every perspective. To people on foot or bike, it’s a hot, loud, dangerous traffic sewer. To people in cars, it’s a congested nightmare every rush hour. Thankfully, there’s enough space on the road to make the first problem a little less bad and fix the second problem.*

The biggest change to Drake would be a third lane heading south from Marin Catholic to 101 and, in one design, north from 101 to El Portal by narrowing lanes from between 15 and 21 feet to a standard 11 feet. This will increase capacity by about 50 percent along the most congestion stretches of the road, a huge boost for drivers that have complained about for years.

Other changes would be squaring off intersections in a few sections to make it safer to walk, new shoulders – formerly Class II bike lanes – and wider sidewalks [2]. Given that there are schools along the route, these are victories for parents and children along the way.

So what’s the problem?

According to an email from MAD [3], and to comments from anti-urbanists from the last time I tackled this problem in November [4], the primary complaint is that narrowing lanes will reduce capacity by slowing traffic. They’ve labelled it a traffic calming exercise, designed to support safety rather than traffic flow.

Even if we look past the morally reprehensible attitude that the safety of people walking isn’t worth protecting, it’s utterly illogical. MAD doesn’t understand how roads work.

A short primer on road capacity functions is in the notes, if you’re interested, but here’s the short of it: moving from a 12-foot lane to an 11-foot lane will decrease the free-flow speed of a road by about 3 percent, but adding a whole new lane will increase capacity much more than what’s lost by that very slight depression in free-flow speed. Add it all up, and the project should boost capacity by 45.5 percent.

But what about right-turn lanes lost? There is still more road capacity, so while the far right lanes might be slower during off-peak times, through drivers who won’t be in the far right lanes won’t need to worry.

To put it very simply, the county wants to slow uncongested travelers by about 3 percent to add 50 percent capacity. This seems like the kind of project MAD would support if it was so worried about traffic.

But MAD’s opposition to adding makes so little sense it boggles the mind. Why would an organization that argued any new homes will cause congestion fight against measures that might actually reduce congestion? The political answer might be the easiest.

It’s an election season, and that Fairfax email was rife with disparaging words for incumbent supervisor Katie Rice and glowing words for the conservative, Kevin Haroff, who has come out against the project [5]. By painting this redesign as a road diet rather than the road widening it is, MAD and its fellow organizations (CVP and Citizen Marin) can say that Rice has no solutions and is beholden to the madness of us urbanists.

If it’s about politics, then MAD is lying about the project to help their candidate and Haroff is complicit in the deception.

Of course, lots of anti-urbanists see a conspiracy to destroy Marin’s character. They probably actually believe that adding bike lanes is just part of that conspiracy. Al Dugan thinks I work for an anti-Marin lobbying group in DC, for instance.

So some people are playing the political game and don’t care if they’re on the technically correct side of a given issue or not as long as their candidate wins. Others want their candidate to win because they think only by cleaning house can they halt the spread of dangerous ideas in Marin. It’s a potent mix, and it‘s leading Marin down a dark path.

If even a project that will boost traffic capacity by nearly 50 percent is successfully painted as a congestion-causing project, there’s something seriously wrong with our politics. When we can’t even agree on what’s real or not, we cannot have a successful government. Fairfax went through this during Frank Egger’s years on the council. San Anselmo is going through its own turmoil with Ford Greene. Sausalito and Marinwood are going through phases when the whole governing body is dominated by people who take this sort of confrontational and personal approach to governing.

The Drake project has its problems, and I’ve highlighted them before, but as a symbol of our political dysfunction it is extremely worrying.


* For today, I’m not going to get into induced demand and the Fundamental Law of Traffic Congestion, which Connor Jones went over this past winter [6]. That’s a deeper problem, but fixing it is something nobody has the political stomach to take on even in San Francisco, let alone Marin.

Level of Service diagram, click to enlarge. Image from Wali Mamon.

** Traffic capacity – how many vehicles a road can carry in an hour – is a function of a road’s design speed and the road’s traffic jam density, or how many vehicles fit per mile when traffic speeds are basically zero:

Maximum Capacity = Number of Lanes × ((Free-flow Speed×Jam Density)/4)

This is called the Greenshields model, which is a reasonable tool for analyzing roads like this. Among other things, it tells us that as speed decreases, the number of cars the road can move per hour actually increases, at least up to a point. This is thanks to the fact that the space between cars decreases as speed decreases, allowing the road to be used more efficiently. Generally speaking, this is around Level of Service (LOS) grade E, though engineers try to keep LOS at around C or D to ensure some slack in the system [7].

According to this model, reducing the free-flow speed by 3 percent in this model to add a new lane will increase the maximum capacity by roughly 45.5 percent.

Works Cited

[1] Marin Against Density, MAD Logo, Digital Image, n.d.

[2] Kentfield Planning Advisory Board, “Sir Francis Drake Boulevard Corridor Rehabilitation” (Kentfield, CA, October 28, 2015).

[3] Marin Against Density, “Attend June 1 ‘Open House’ -- Forward This to Friends.,” June 1, 2016.

[4] David Edmondson, “Build Something Better on South Sir Francis Drake,” The Greater Marin, November 18, 2015.

[5] Issues,” Kevin Haroff for Supervisor District 2, accessed June 2, 2016.

[6] Connor Jones, “The Street Economics of Induced Demand,The Greater Marin, December 21, 2015; Connor Jones, “The Four Biggest Myths about Induced Demand,” The Greater Marin, January 4, 2016.

[7] Francis Vanek et al., Sustainable Transportation Systems Engineering: Evaluation & Implementation (New York: McGraw-Hill Education, 2014), chap. 7; Wali Memon, “Highway Level of Service,” October 12, 2012.

What if the Bay Area had never lost its railroads?

With the Northwestern Pacific reopening up soon as a limited commuter rail service - even though it need not - one wonders what the Bay Area would have looked like had it reinvested in its rail lines over the years rather than just rebuilt them. Well, it probably would look something like this:

Map by Theo Ditsek. Click to go to his post, where you can find a full-size image.

Map by Theo Ditsek. Click to go to his post, where you can find a full-size image.

I am always pleased to see our region's transit system reimagined by others, but I'm especially happy to see Theo Ditsek, one of the internet's more prolific transit map hobbyists, tackle the Bay Area, and tackle it like he did. From what I can see, he didn't leave a single rail line unused: there are rail lines from West Marin to Sausalito, subways through San Francisco, commuter rail to Santa Cruz, and shuttle lines heading every which way. My favorite, though, is the railroad to Tahoe.

Perhaps my only complaint is the lack of rail links from Marin and Vallejo to San Francisco, both for commuters and long-distance travelers. Beyond this, however, it's a truly marvelous piece of work. Now if only I could get it as a poster...

SMART Train alcohol policy comes up short

SMART wants to limit alcohol on its trains to only what patrons buy at the concessionaire. Experience from Caltrain shows that allowing riders to BYOB policy is not just good policy, but helps embed the system further into the culture of riders. 

Caltrain riders hanging out. Image from SF2G.com

Caltrain riders hanging out. Image from SF2G.com

For 18 years I have ridden Caltrain - and in that time I have brought aboard and consumed hundreds of bottles and cans of beers, which I have enjoyed legally on the train. This includes frequent patronage of the semi-official Party Car formed by the cyclists on Caltrain.

Starting in 2000, alcohol consumption on Caltrain increased exponentially with the opening of AT&T park, home of the San Francisco Giants. Giants fans have flocked to the train, riding up the Peninsula with cases of beer and bottles of who knows what, safely being carried to and from the games. At some point, Caltrain decided to ban alcohol on trains running after 9 PM only IF there is an event - primarily Giants games but also Sharks games, concerts at AT&T Park, and now 49ers games and concerts at Levi's Stadium. That late, the consumption before and during the events reach enough of a pitch that it was prudent to put a limit on the policy. Over the years the train has also served hugely alcohol-fueled events like Bay to Breakers, Hardly Strictly Bluegrass, Octoberfests, etc... frequently on the same day.

Generally speaking, this open BYOB policy on Caltrain has been a success. Problems are generally very rare, given the nature of the train as primarily a commuter rail with a higher level fare. It has been an attraction to the train that has a nominal positive influence on overall ridership numbers.

At the end of 2016, I will start riding a new train line - the SMART train in Sonoma County. It will function as primarily commuter rail, running almost exclusively during peak commute hours, with a fare structure prohibitive to general miscreants, making it nominally similar to Caltrain, except that it will serve no special events like the Giants as there are no major sports or entertainment venues on the train line.

SMART has released their draft code of conduct for the train. It includes a policy of NO BYOB. They don't have a no alcohol policy because they have an operating theorem of having a bar car on the train. I am very disappointed by this policy - I find it rider hostile and that it will have a negative impact on the rider experience and overall ridership. While there will be a bar car - there are bar cars on the Amtrak Capitol Corridor too, but in a place like Sonoma County with an excellent selection of beer and wine, to limit riders to the meager selection of a bar car is misguided.

There is, of course, the suspicion that the rationale is not to prevent unruly behavior, but to support whatever vendor they get for their bar car. This is understandable, for the most part because the decision to put a bar car onto the trains is misguided at best. SMART put out a presentation on the bar car where they are toying with giving free rent to the vendor for a return of a percentage of the profits. This is predicated on the presumption that a bar car will turn a profit - my experience from the Amtrak Capitols is that the bar car is at best a loss leader, not a profit center. This is especially true when you consider that SMART has wasted valuable train space to build the bar car.

I personally think that the potential ridership of SMART should make it clear that the no BYOB policy is not in line with the overall goals of the train. They should reconsider this path and allow riders to bring their own beverages onto the train.

If you agree - please email SMART at info@sonomamarintrain.org.

Originally Published: Murphy, John. 2016. “SMART Train Alcohol Policy.” Holier than You Blog. May 18.

Dear Readers

I’ve been gone. Sorry about that.

After finishing up a tough second semester at Cornell University studying for a master’s in City and Regional Planning, I’m off to Manila, capital of the Philippines, for an internship with the country's Department of Transportation, so expect some dispatches from there.

On the upside, having a summer off means I’ll have more time to devote to Marin’s issues specifically, like the aftermath of the election (seriously, Citizen Marin/CVP/Marin Post? The best conservative you could find is Al Dugan?).

So what have been the big takeaways from my first year as a master’s student? First, that 11-foot wide lanes really are just fine for putting traffic through Sir Francis Drake Avenue. That a freeway, or a road, has the highest capacity at LOS D or E. That the only way to solve traffic congestion is with pricing.

That the Rawlsian ethical imperative for equity and a market urbanist philosophy actually can mix, and that the Bay Area is failing at both miserably.

I also found evidence, through a statistical analysis of Census data, that adding more homes to a given neighborhood allows it to hold onto its poor and middle-class populations better. So if you care about keeping gentrification at bay, you should build new homes, luxury and affordable alike.

And that one of the reasons the trains of the early 20th century aren’t really around anymore is because they were so damn slow, like less than 30mph slow, and the companies didn’t realize the point of transportation is transportation, not luxury.

Speaking of old trains, in between classes I finished up the Map of the Mid-Atlantic, a Cleveland frequent bus map, and started work on a Saint Louis of 1921 map.

I have a couple of papers to adapt to blog posts here, including that statistics analysis, and also some that I promised long ago that still have not been published. Given that on Thursday I’ll be flying for more than 19 hours, I will hopefully get some stuff done for once.

 - David

Progress on the DC & Baltimore Map

It’s been a while, clearly, but there is progress. I’m in final check-up and labelling, which is a long slog. Pennsylvania, Western Maryland, Maryland & Pennsylvania, and the ferries are done. I’m also finding out whether or not I missed some lines, and boy did I miss a biggie. First, the progress map:


I had to re-do a couple of the river ferry lines to fit in the York River line. Mobjack Bay, which had ferry service to Norfolk, had to get cut. That’s fine, given the limitations of the map and the relative importance of getting to West Point from Baltimore compared to Mobjack service, but it’s a loss nonetheless.

I’ve also finished up the Pocomoke & Occohannock line, which was a true pain, with different service in either direction for almost every day of the week: 11 patterns in 7 days. Ugh.

Now, I’m on the final check-up: label all the lines coherently, color the lines appropriately, make sure there aren’t any spelling errors or misplaced lines or circles, and check to ensure all service is included. Labeling is some Excel work, which looks like so:

Percentages are the percent of times a train lumped into a given line will stop at that stop. It takes a lot of futzing to cut this down.

Alas that last thing, checking to ensure all service, is the Big Deal. 

When I started the map, I was only going to do lines that started within the confines of the contemporary DC & Baltimore Metropolitan Area. As I went along, I decided it would be more important to show all the connections there is space for. As a result, there are a few lines I missed. Hagerstown & Fredericksburg, for instance, has the Northern and Williamsport divisions, which I’ve now added in, but more important is what happens around Wilmington.

Around Wilmington, I missed the Wilmington Division of the Philadelphia & Reading, the Pomeroy Branch of the Pennsylvania, and the Landenburg Branch of the B&O. These all interconnect with either one another or Pennsy’s Octoraro Branch or both, which means some crazytown geometry and a reworking of some of the first lines I drew. How fun. On the plus side, it means we’ll get to see every station in Delaware, so it’ll be the first complete state in this series. Yay! Here’s the geographic map of what I’m up to:

Roughly the area bounded by the red box: Chadd's Ford Junction to Avondale to Newark to Wilmington. The electric lines shown here aren't listed in the Official Guide to the Railways, so they don't get to be included.

I also found a whole new railroad that I hadn’t seen, the New Park & Fawn Grove:

So yes, folks, I’m plugging away. What’s done:

  • Reviewed and labelled the Pennsylvania; Washington, Bandywine & Point Lookout; Maryland & Pennsylvania; Hagerstown & Frederick; Chesapeake Beach; Washington, Baltimore & Annapolis; Maryland, Delaware & Virginia; Western Maryland; and ferry lines
  • Added the Stewartstown and New Park & Fawn Grove railroads
  • The coastline

Here’s the to-do list in no particular order:

  • Add a couple of barrier island inlets on the Atlantic seaboard of the Delmarva Peninsula
  • Finish up the Wilmington Division of the Philadelphia & Reading, the Pomeroy Branch of the Pennsylvania, and the Landenburg Branch of the B&O.
  • Review the B&O; C&O; Norfolk & Western; Potomac, Fredericksburg & Piedmont; RF&P; Southern; Washington & Old Dominion; and Philadelphia & Reading railroads
  • Convert white label outlines to 80% transparent
  • Fill in the western Virginia whitespace
  • Add legend, list of service, and map title

Pre-orders are available in the Map Store.

The four biggest myths about induced demand

This post, by Connor Jones, originally appeared on Urbanist.co. It has been edited to include citations, including the correction of dead links.

Two weeks ago, I laid out the economic argument for induced demand (Jones 2014b): the idea that building more roads does not reduce congestion. It is a simple model that uses concepts from Economics 101 to explain the relationship between road construction and driving behavior.

Even so, this idea, like many associated with new urbanism, challenges the status quo. As such, there’s pushback. To ensure that no falsehoods go unchallenged, I decided to examine the claims in two articles that seek to discredit induced demand as a property. The first is a blog post from the Cato Institute (O’Toole 2014) written in response to a Wired article on the subject (Mann 2014) published last month, and the second is a Weekly Standard (Last 2011) story written three years ago (which is still fresh in urban planning time). Here are the four most prominent false assertions upon which the articles rely.

1. Since roadway capacity is not the only factor affecting driving, induced demand is a flawed model.

This misrepresentation was trotted out by the Cato Institute, which attempted to discredit the academic research of Gilles Duranton of the University of Pennsylvania and Matthew A. Turner of the University of Toronto, who measured the elasticity of demand of vehicle miles traveled (2011).*

On average, driving grew more than twice as fast as lane miles. But in Boston between 1983 and 1993, freeway capacities grew by less than 1 percent, while driving grew by more than 35 percent. In Madison, capacities grew by 35 percent, while driving grew by less than 20 percent. The wide range in differences between urban areas suggests that, not only are Duranton & Turner’s elasticities wrong, their standard errors are far too low. (O'Toole 2014)

Duranton & Turner’s headline finding was that the elasticity of demand in the transportation market is 1, according to roadway data from 1980 to 2000. In other words, holding other factors constant, a 20 percent increase in roadway miles elicits a 20 percent increase in vehicle miles traveled. “We found that there’s this perfect one-to-one relationship,” Turner told Wired (Mann 2014).

Cato fails to account the other variables that affect driving patterns like geography, population growth, and socioeconomic characteristics that Duranton & Turner specifically control for. Simply noting that all cities’ freeway capacities and driving patterns don’t fluctuate in lock step does not show anything.

Duranton & Turner sought to find the relationship between two variables alone and found a striking relationship. We live in the real world, and there are other factors that affect people’s behavior.

2. Cities that have invested in public interstates have seen long-term reduction in congestion.

The Weekly Standard blithely throws out this claim without qualifying it in any way:

The Texas Transportation Institute’s annual Mobility Report, for instance, demonstrates an uncanny correlation between capacity and traffic congestion: Areas that add capacity tend to have lower levels of congestion. (Last 2011)

First off, that’s not what the the authors of the Texas Transportation Institute’s Mobility Report (Schrank, Eisele, and Lomax 2012) found. Instead, they wrote that “additional roadways reduce the rate of congestion increase,” which is a substantively different assertion.

Additionally, their analysis is based on the assumption that roadway growth (supply) and vehicle miles traveled (quantity demanded) are independent of each other. While there are certainly other factors involved, the built environment contributes significantly to people’s behavior. Ignoring this fact is tantamount to building a new road, observing an increase in vehicle miles traveled, then assuming it would have happened anyway. This methodology leads to a skewed result, which isn’t matched by other studies.

The most robust study on the relationship between congestion and roadway growth comes from the Victoria Transport Policy Institute, which found that “Traffic congestion tends to maintain equilibrium. Congestion reaches a point at which it constrains further growth in peak-period trips. If road capacity increases, the number of peak-period trips also increases until congestion again limits further traffic growth.” (Littman 2015)

Plenty of academically-minded people before me have established the economic model. For one, Douglass B. Lee Jr. at the World Bank provides a more rigorous explanation (Lee, n.d.). A meta-analysis of induced demand studies by Robert Cervero in 2003** (Cervero 2003) found strong evidence of the existence of the phenomena, though different researchers have established different elasticity quotients . Recently, Duranton & Turner derived an elasticity of 1 with a very low standard error (2011).

(The Texas Transportation Institute study has several other problems that Tanya Snyder at Streetsblog USA (2013) and Todd Litman at the Transport Policy Institute (2014) can address more thoroughly than I can.)

Interstates according to writers at the Cato Institute and the Weekly Standard. Image by scot63us, on Flickr (2010).

3. Automotive transportation is the most efficient way of moving people around a city.

This contention isn’t really even a myth—it’s a fabrication. It has no basis in reality. This point was appended to the end of the Weekly Standard article:

A metropolitan area typically has about half as many jobs as people. But, because of geographical constraints, not every job is accessible to every person. Highways are, far and away, the most efficient way of delivering people to a job. (Last 2011)

In most American cities, auto transportation is the best readily-available way to transport people because there are no other options. That does not in any way imply that it’s the most efficient way to organize a city. On the contrary, car dependence is both inefficient and wasteful:

  • University of Michigan study: “Overall, in 2010, BTU per person mile was 4,218 for driving versus 2,691 for flying. Other modes of transportation: Amtrak trains (1,668), motorcycles (2,675) and transit buses (3,347).” (Sivak 2015)***
  • Portland’s dense development patterns yields $2.6 billion in yearly savings, which amounts to a 3.0 percent income bump relative to the average citizen of the United States (Jones 2014a).
  • Automobile congestion as a result of publicly-subsidized highways costs Americans at least $45 billion every year (Jaffe 2013).

4. Vehicle miles traveled isn’t an important metric.

This is a confused contention that doesn’t hold up to any scrutiny. For some reason, we shouldn’t be focusing on vehicle miles traveled as a metric because… we don’t like doing it? Again, from TWS:

[Principal at Demographia Wendell Cox] maintains that the entire framing of the issue is faulty: “Latent demand” for a highway, he notes, isn’t actually a desire to drive on that stretch of road. People only want the road as a means to an end. “Transportation is not a primary activity,” Cox explains. “There is no ‘love affair’ with the automobile. Driving is not something we would choose to do.” […] In other words, a metric like “vehicle-miles traveled” is only superficially important. (Last 2011)

Plenty of economic goods are means to an end. No client wants to pay up to mount a legal defense, but they do it anyway because they have to. Just because it’s a means to an end doesn’t mean we have to spend the money for it. With dense development and healthy public transit, families are able to spend less time in traffic and fewer dollars on gas without sacrificing mobility. Maximizing vehicle miles traveled should not be anyone’s objective.

Induced demand is an economic property with solid evidence

The key insight from the market model is that increasing roadway capacity will only make sprawl worse and won’t fight congestion. While car dependence hurts public health and wastes money, this economic principle does not imply that all highway construction is misguided. All planning is local. (Like politics.) There are plenty of good highway projects, but they must be balanced with investment in transit so that our cities can be strong, diverse communities where having a car isn’t a prerequisite for full citizenship.

End notes

For a more intuitive explanation of induced demand, see this insightful post from Greater Greater Washington on how building public roads to fill “latent demand” is like putting out more and more free hamburgers to feed people (Alpert 2012).

*Elasticity of demand (Heakal 2015) is the percent change in quantity divided percent change in price that measures responsive consumers are to changing their behavior given a price increase. This quantity can be visualized by the slope of the demand line.

**Editor's Note: The Cervero piece from 2001 originally cited by Jones has disappeared, but Cervero released an update to his research from 2003 that reinforces the points made in that earlier study. We have linked to that later study and updated the text to avoid confusion.

***Editor's Note: The Sivak piece from 2014 originally cited by Jones has disappeared, but, like Cervero, Sivak released an update to his research that reinforces the points made in that earlier study. We have linked to that later study.

Works Cited

In keeping with my university's standards, future blog posts will use in-text citations and a works cited. Often, these will be behind a paywall; please email me at thegreatermarin@gmail.com if you would like the full text.

Alpert, David. 2012. “Comment of the Week: Induced Demand Is Free Fast Food.” Greater Greater Washington. September 4. http://greatergreaterwashington.org/post/16029/induced-demand-is-like-free-fast-food/.

Cervero, Robert. 2003. “Road Expansion, Urban Growth, and Induced Travel: A Path Analysis.” Journal of the American Planning Association 69 (2): 145–63. doi:10.1080/01944360308976303.

Duranton, Gilles, and Matthew A. Turner. 2011. “The Fundamental Law of Road Congestion: Evidence from US Cities.” The American Economic Review 101 (6): 2616–52.

Heakal, Reem. 2015. “Economics Basics: Elasticity.” Investopedia. Accessed December 17. http://www.investopedia.com/university/economics/economics4.asp.

Jaffe, Eric. 2013. “The U.S. Transportation System Has $100 Billion Worth of Inefficiencies.” CityLab, October 1. http://www.theatlanticcities.com/jobs-and-economy/2013/10/us-transportation-system-has-100-billion-worth-inefficiencies/7076/.

Jones, Connor. 2014a. “Want to Reduce Reliance on Foreign Oil? Start with Walkability.” Urbanist.co. June 12. http://urbanist.co/want-reduce-reliance-foreign-oil-start-walkability/.

———. 2014b. “The Street Economics of Induced Demand.” Urbanist.co. June 25. http://urbanist.co/street-economics-induced-demand/.

Last, Jonathan V. 2011. “More Highways, Less Congestion.” Weekly Standard, March 7.

Lee, Douglas B. Jr. n.d. “Appendix B: Induced Traffic and Induced Demand.” In Concepts of Induced Demand. Washington, DC: World Bank.

Littman, Todd. 2014. “Congestion Costing Critique: Critical Evaluation of the ‘Urban Mobility Report.’” Victoria, BC: Victoria Transportation Policy Institute. http://www.vtpi.org/UMR_critique.pdf.

———. 2015. “Generated Traffic and Induced Travel: Implications for Transport Planning.” Victoria, BC: Victoria Transportation Policy Institute. http://www.vtpi.org/gentraf.pdf.

Mann, Adam. 2014. “What’s Up With That: Building Bigger Roads Actually Makes Traffic Worse.” Wired, June 17. http://www.wired.com/2014/06/wuwt-traffic-induced-demand/.

O’Toole, Randal. 2014. “Debunking the Induced-Demand Myth.” Cato at Liberty. June 18. http://www.cato.org/blog/debunking-induced-demand-myth.

Schrank, David, Bill Eisele, and Tim Lomax. 2012. “2012 Urban Mobility Report.” Urban Mobility Report. College Station, TX: Texas A&M Transportation Institute. http://d2dtl5nnlpfr0r.cloudfront.net/tti.tamu.edu/documents/ums/archive/mobility-report-2012.pdf.

scot63us. 2010. Highway. Digital Photograph. Flickr. https://www.flickr.com/photos/scottummy/4971054099/.

Sivak, Michael. 2015. “Energy Intensities of Flying and Driving.” UMTRI-2015-14. Ann Arbor, MI: University of Michigan. http://www.umtri.umich.edu/what-were-doing/news/planes-trains-and-automobiles-traveling-car-uses-most-energy.

Snyder, Tanya. 2013. “TTI Urban Mobility Report Bungles Congestion Analysis Yet Again.” Streetsblog USA. February 5. http://usa.streetsblog.org/2013/02/05/tti-urban-mobility-report-bungles-congestion-analysis-yet-again/.

Original Post

Jones, Connor. 2014c. “The Four Biggest Myths about Induced Demand.” Urbanist.co. July 7. http://urbanist.co/busting-four-biggest-myths-induced-demand/.

Connor Jones

Jones graduated Georgetown University with a degree in Mathematics and Economics in 2014, after having worked part-time for a several years covering local planning projects in Washington, DC and Atlanta, Ga. for alternative newsweeklies.

The street economics of induced demand

This post, by Connor Jones, originally appeared on Urbanist.co. In keeping with The Greater Marin's style, in-text citations have been added.

There’s a lot of misinformation out there about induced demand (Littman 2015), a widely established city planning model that attempts to explain why cities tend to maintain a steady state of congestion. I’ll go into some more detail on the theory of induced demand later, but I wanted to start with the economic model.

Image by Connor Jones, Urbanist.co

Auto transportation market

The amount that people drive is determined by market forces of supply and demand. Consumer preferences about driving are provided by the demand function, which shows that drivers choose to drive less as the cost of driving increases. Conversely, the supply function is a representation of the existing stock of roadways. In this case, the cost of auto travel is congestion.

Model Assumptions

The principle assumption is that the primary variable cost of auto travel is the time it takes to get from point A to point B, which is reasonable since the fixed cost of owning a car far outweighs the cost of gas for a majority of consumers. (Consider a worker who makes $15 an hour full time. An increase in average commute times by 12 minutes per day decreases his income by $750 a year—the equivalent of a 2.5 percent pay cut.) Even so, the price of gas does factor into people’s driving decisions, so we have to assume that the price of gas is constant. Furthermore, we assume that consumers’ driving preferences are constant. (More on this assumption later.)


The most common justification cited for building a new roadway is to reduce congestion, which makes sense. Most localities will commission a traffic study, which frequently assume constant growth of drivers and therefore also congestion.

Induced demand shows a different story, however. According to our model, the increase in the stock of roadways shifts the supply curve out, which does, in fact, reduce congestion in the short term, but, not as much as it would were there not growth in the number of vehicle miles traveled. Since the cost of driving is reduced, drivers both make longer and more frequent trips in the short term. As we can see from our graph, the number of vehicle miles traveled increases from q1 to q2.

In the long term, the reduced congestion encourages the construction of less dense housing developments far from the city center. With the new roadway, commuters can live farther from their places of work and leisure at the same cost. Over time, these developments shift the demand curve out, reducing the gains against congestion and further increasing the number of vehicle miles traveled.

The effect on cities

The auto transportation market explains the intuition that building more freeways makes a city more car-dependent and encourages sprawl. In very many documented cases (which I will summarize on Friday), building a new roadway does not reduce congestion for very long. Cities that invest in auto infrastructure do not see improvement in congestion (Gehl 2010).

What traffic engineers assume

According to Jeff Speck in Walkable City (which I’m still reading), traffic engineers commonly assume that demand for roads will increase at a constant rate, year over year (2013). With that assumption in place, you can see how they could come to the conclusion that a new freeway will improve congestion in the long term. The market movement without induced demand is illustrated below.

Image by Connor Jones, Urbanist.co

The conclusions reached are vastly different. Under these assumptions, after opening the freeway, there is no increase in vehicle miles traveled (which is not corroborated by real-world data.) The reduction of congestion is larger than the induced demand model predicts, and the only growth in demand is independent on roadway growth.


The relative sizes of the movements along the curves will vary depending on the slope of the demand curve (which is dependent upon consumer preferences, which vary from place to place). We will examine the variation in the relative effects of induced demand later this week.

Public policy implications

Government-supported roads are effectively subsidies for motorists. Without as much investment in roads, the free market would have a greater incentive to create public transportation and dense housing options in city centers. For this reason, the enemy of the walkable city is the six-lane freeway. As I have argued before (and will no doubt argue again), walkability serves several public policy goals at once (Jones 2014a).

Works Cited

h/t to Jeff Speck’s Walkable City, where I found much of this information. All resources:

Federal Highway Administration. 2012. “Induced Travel: Frequently Asked Questions.” Office of Planning, Environment, & Realty (HEP). December 3. http://www.fhwa.dot.gov/planning/itfaq.cfm#q4.

Gehl, Jan. 2010. Cities for People. Washington, DC: Island Press.

Jones, Connor. 2014a. “Want to Reduce Reliance on Foreign Oil? Start with Walkability.” Urbanist.co. June 12. http://urbanist.co/want-reduce-reliance-foreign-oil-start-walkability/.

Littman, Todd. 2015. “Generated Traffic and Induced Travel: Implications for Transport Planning.” Victoria, BC: Victoria Transportation Policy Institute. http://www.vtpi.org/gentraf.pdf.

Mann, Adam. 2014. “What’s Up With That: Building Bigger Roads Actually Makes Traffic Worse.” Wired, June 17. http://www.wired.com/2014/06/wuwt-traffic-induced-demand/.

Schmitt, Angie. 2012. “Report: Traffic Studies Systematically Overstate Benefits of Road Projects.” Streetsblog USA. July 6. http://usa.streetsblog.org/2012/07/06/report-traffic-studies-systematically-overstate-the-benefits-of-road-projects/.

Speck, Jeff. 2013. Walkable City: How Downtown Can Save America, One Step at a Time. 1st edition. New York: North Point Press.

Original Post

Jones, Connor. 2014b. “The Street Economics of Induced Demand.” Urbanist.co. June 25. http://urbanist.co/street-economics-induced-demand/. Reposted with permission.