My 2016 Politics Posts in One Place

To aid those who will send me to the re-education camps, I list my 2016 politics (not transport politics) posts in one place:

On the ‘need’ for road funding.

Lots of numbers are thrown around about “need” for road funding.

For instance, in Fix It First, (2011)   we said a “National Cooperative Highway Research Program [“NCHRP Project 20-24(49): Future financing options to meet highway and transit needs.” Table A1] report finds an annual need of $188.4 billion in 2007 dollars to maintain existing highway infrastructure, of which $109.8 billion is capital and the remainder ($78.6 billion ) is operations and maintenance costs (National Cooperative Highway Research Program [NCHRP] 2006).”

The National Surface Transportation Policy and Revenue Study Commission also has estimates which are higher, and the ASCE Report Card says about surface transportation (roads, bridges, and transit) “We are facing a funding gap of about $94 billion a year with our current spending levels.” These numbers are of course different, but all very similar, especially given the large span of years over which they have been reported. There is a reason they are similar, they derive from the same source.

Minetta Lane, New York.
Minetta Lane, New York.

Now where do the numbers come from? The root source of most of these “need” numbers are various runs of the Federal Highway Administrations HERS (HERS-ST) system (described here), which has been evaluated by GAO (here and here). HERS  is probably the best available system for building up an “engineering” estimate of needs, based on the condition of individual links in their database. While it is the best available, HERS is far from perfect. Nothing is perfect. As the description says

The program is noteworthy for its ability to conveniently perform sensitivity analyses in order to test whether the solution of recommended investments is robust to changing system goals and underlying parameters. Note that HERS does not reallocate traffic to reflect highway improvements. However, traffic growth induced by improved capacity and operating conditions is included, with half the estimated user benefits counted for induced traffic, consistent with consumer surplus principles. (FHWA, 2002d).

As with anything, the output depends on the inputs. If traffic is expected to grow, the need for additional capacity resulting from this kind of comparative statics analysis will be higher than if traffic is expected to be flat. The forecasts are exogenous, this is not a transportation planning model.  And while “induced demand” is sort of included, unless it is analyzing network flows, it can’t really be. If a link upstream of the link in question is improved, it will induce additional traffic both upstream and downstream (and reduce traffic on competing links). This induced traffic is due to rerouting, switching time of day or day of week, switching modes, making longer trips, and making trips that would not otherwise have been made, among other sources, and over the long term by encouraging new development. A network analysis is required to assess how much this is.

HERS is far better for assessing pavement  quality issues (for which links are separable) than level of service issues (for which they are not). To be fair, the more recent analysis conducted for ASCE  tries to address (p.34) this question in part by looking at a national Freight Analysis Framework network, and assigning (and reassigning traffic) to that, but that is a really crude sketch network, and the equilibration assumptions are not clearly laid out, and that only looks at choice of route which will be used, not the level of demand itself, and how those interact.

So using the HERS estimates for preservation is sounder than using the estimate for ensuring the same Level of Service. These two numbers are often combined and conflated. They should not be.

There isalso  undoubtedly some arbitrariness in pavement performance standards, bridge conditions, and other infrastructures as well, though I doubt it rises to the level of arbitrariness of highway LOS.

When being a responsible consumer of such numbers, ensure they are presented clearly. Estimates of need are likely overstated, both for reasons of methodology and due to motivated reasoning.

There are “needs” to make sure the highway system does not eventually crumble into dust, and there are “needs” to ensure it maintains the same level of service. One of these needs is not like the other.

Could a Trump administration kill the Green Line extension? | Boston Globe

I was interviewed by Nicole Dungca at the Boston Globe who writes “Could a Trump administration kill the Green Line extension?”  (NOVEMBER 18, 2016). My quotes in context below:

1. Sanctuary Cities

Somerville and Cambridge, both involved in the Green Line extension, are so-called “sanctuary cities,” which generally do not cooperate with immigration officials to detain undocumented immigrants. A strong critic of undocumented immigrants during his campaign, Trump made headlines promising an end to federal funding to such cities.

Would moving on those promises cut funding to Green Line extension project, which is counting on that massive federal grant?

If local and national transportation funding watchers are worried, they’re not letting on. University of Minnesota’s David M. Levinson, a transportation analyst and professor, said that defunding sanctuary cities could affect so many communities that it’s essentially a non-starter. And several leaders of sanctuary cities have already said they plan to be defiant if Trump does make good on his promise.

2. Walking back commitments:

“In terms of spending priorities, most money has been allocated from the previous transportation bill,” Levinson said. “The question is whether there’s new money coming up in anything.”

Trump has touted a $3 trillion spending plan that relies mostly on giving incentives to convince private firms to invest in building roads, bridges, and other big transportation projects — but Levinson said that it would likely only apply to select projects with very high returns. And that description doesn’t generally apply to the money-losing public transit projects that urban centers need to battle decades of underinvestment.

3. New spending:

What if Trump actually wants to make extending transit lines or spending big federal dollars on transportation a priority?

He’ll face a skeptical Congress. With Republicans capturing the majority in both chambers, any ambitions for substantially increasing federal funding for transit would likely face a lot of pushback unless it relies on substantial private sector help.

Many transportation watchers are taking a “wait and see” attitude. Pollack said she’ll keep talking with the national American Public Transportation Association, and the American Association of State Highway and Transportation Officials to keep updated on whether there will be any changes in funding.

And as news trickles out about Trump’s new administration in the coming weeks, transportation is likely to take a back seat to talks about higher-profile appointments, such as secretary of state and attorney general.

“There’s a lot of uncertainty about this administration,” said Levinson. “But transportation spending doesn’t seem to be the thing they should be worried about in this administration.”

On ASCE’s Report Card

I was interviewed by Angie Schmitt at Streetblog on Infrastructure propaganda. Her article with Ben Fried: Chuck Schumer and Nancy Pelosi Are Falling for Infrastructure Propaganda discusses the issue of the American Society of Civil Engineer’s (ASCE) Report Card, with choice quotes from Strong Towns’ Chuck Marohn (also a former student). Quoting:

David Levinson, a civil engineering professor at the University of Minnesota, agrees there is a conflict. “I just went on radio and said ASCE had ‘motivated reasoning,’ so yeah, I think there is a conflict of interest,” he told Streetsblog via email. “Now that doesn’t mean they are wrong, necessarily, but that they do have the incentive to be inflated.”

For an example of the values that inform the ASCE’s report card, look at the section on highways.

The report gives the nation’s roads a “D” letter grade. Of the total $170 billion in annual road spending ACSE calls for, half is for the repair and operation of existing infrastructure. The other half is for expanding highways.

In making the case for highway expansion, ASCE draws on sources like the Texas Transportation Institute’s famously flawed congestion index, which tends to penalize places that shorten the distances people have to travel, while rewarding places that increase those distances. In other words, it is biased toward sprawl and against walkable places.

As Levinson points out, congestion could be managed by properly pricing roads, rather than expanding them, but this option is completely ignored by ASCE.

Some infrastructure in Washington DC.  Pavements are beginning to fail, yet cars continue to drive.
Some infrastructure in Washington DC. Pavements are beginning to fail, yet cars continue to drive.

Are roads and water both a “D”?

In the ASCE report card drinking water gets a “D”. They write:

“At the dawn of the 21st century, much of our drinking water infrastructure is nearing the end of its useful life. There are an estimated 240,000 water main breaks per year in the United States. Assuming every pipe would need to be replaced, the cost over the coming decades could reach more than $1 trillion, according to the American Water Works Association (AWWA). The quality of drinking water in the United States remains universally high, however. Even though pipes and mains are frequently more than 100 years old and in need of replacement, outbreaks of disease attributable to drinking water are rare.”

To be sure 240,000 sounds like a big number, but it is less than 1 break per 1000 people in the US (I know water mains serve more than 1 person, but that count of failures include many, many small ones for each one that gets on the news). Now I am sure drinking water in the US could be better somehow, and Flint is a tragedy, but almost everyone in the US turns on the tap, gets water instantly. The water is clean, and so almost no one in the US dies from drinking water anymore. A 90+/100 usually scores an A. The Environmental Engineers are to be commended. They have been so good, they almost put themselves out of a job.

A Political Economy of Access: Infrastructure, Networks, Cities, and Institutions by David M. Levinson and David A. King
A Political Economy of Access: Infrastructure, Networks, Cities, and Institutions by David M. Levinson and David A. King

So let’s grant that over the next 100 years all the pipes need to be replaced due to potential failures (broken mains, leakages, etc.). And let’s grant that would be $1T (ignoring inflation/discounting etc) (I have no clue what the real number is). This seems a lot, like about $3000/person. But that’s still only $10B/year (or $30/person/year) which is $2.50 per person per monthly water bill for capital replacement, or $0.10/day/person, which seems eminently doable, and is the responsibility of local water utilities, either private regulated utilities, and/or owned by local governments.

If you think it needs to be done in 50 years, double it. This hardly constitutes a crisis. In medical terms, the condition is chronic, not acute.

About Roads, they write:

“Forty-two percent of America’s major urban highways remain congested, costing the economy an estimated $101 billion in wasted time and fuel annually. While the conditions have improved in the near term, and Federal, state, and local capital investments for road infrastructure increased to $91 billion annually, that level of investment is insufficient and still projected to result in a decline in conditions and performance in the long term. Currently, the Federal Highway Administration estimates that $170 billion in capital investment would be needed on an annual basis to significantly improve conditions and performance.”

  • First, note that phrase “conditions AND performance”, that means maintaining highway LOS, not just pavement surface quality.
  • Second, note this is for unpriced roads. If roads were properly priced, congestion would go down significantly. (Pricing also raises revenues).
  • Third, it makes no assumption about more efficient use of roads from Autonomous Vehicles. While they are not deployed yet, over the expected lifetime of new infrastructure, they will be. It takes a particular type of ostrich to just ignore this.
  • Fourth, again they don’t make the case as to why this (or any of it) is a federal rather than state and local responsibility. They do say “federal, state, and local” need to come up with the money, and they are indifferent to which (green is green). But it matters in practice.
  • Fifth $101 Billion in wasted time and fuel is supposed to be solved by $170 Billion in investment. I am not exactly clear on how much of that $170B (or more precisely $170B-$91B= $79B) goes to reduce, but not eliminate, $101B, (it seems like half since $85B is supposed to go to State of Good Repair, so I guess the other $85B goes to safety and congestion issues) from their report)  but it doesn’t sound like it’s a good idea. It’s not at all clear the benefits outweigh the costs, unless that $85B reduces congestion or fuel costs by at least $85B to $16B ($101-$85=$16) which seems unlikely.
  • Sixth $101B sounds like a lot of time and money, this is of course a TTI Urban Mobility Report estimate. It is a little over $300/person/year, or less than $1/day/person, or less than $0.25/trip, which is again annoying and perhaps needlessly wasteful, but not a crisis.

So yeah, Roads get scored a “D”, and this is probably more apt than water, and we could do better, but I do not think the solutions are what they think they are.

So what does a “D” even mean?

The rubric is given at:

The infrastructure is in poor to fair condition and mostly below standard, with many elements approaching the end of their service life. A large portion of the system exhibits significant deterioration. Condition and capacity are of significant concern with strong risk of failure.”

My students would complain if the grading rubric were that vague. “Good” and “Fair” (and so on) conditions are labels for specific performance levels in each infrastructure (roads, bridges, I assume the others). I think they are defined according to specific reasonably objective performance levels for a specific technology (for example pavement roughness for pavement, bridge conditions for bridges).  But how is a good road compared to a good bridge?

But when ASCE says “many”, how many? When they say “a large portion”, what share? When they say “strong risk of failure”, how strong. If a bridge were known to have even a 0.1% chance of failing tomorrow, it would be closed immediately by the Professional Engineer in charge. Failing in a bridge is very different than failing in roads (and pavement condition failure is different from level of service failure).

Not all roads are in poor to fair condition. Many are brand new. Many elements ARE nearing end of life, but of course, many should be, so are many people, that is what is meant about managing across the lifecycle. It would be terribly inefficient if everything both opened the same year and then failed 50 years in the future also in the same year. When we know where a facility is on the lifecycle, at some appropriate point we make a repair/replace/abandon decision.

I can see grading individual elements in a given technology (An “A” vs a “D” bridge is perhaps a meaningful comparison if it represents the same objective elements). I am not sure how they average bridges across the system, (we all know how to do an average, but there are lots of ways to do averages, and they usually should be weighted somehow). You could certainly have an average bridge or average road, but those are always fine, usually people talk about the share of bridges which are structurally deficient (or “functionally obsolete”) or in “poor condition”.  For bridges, see e.g. the report on which this USA Today article is based,  or this similar bridge tool by T4 America.

Also note that scoring a report card grade of “A” (and maybe “B” or “C”) is overbuilding for anything but a brand-new infrastructure.


On Trump’s Infrastructure Plan

Brad Plumer at Vox interviewed me about “Trump’s plan to finance infrastructure by offering tax breaks to developers working on projects with revenue streams (like toll roads).” He wrote this article: Donald Trump’s infrastructure plan wouldn’t actually fix America’s infrastructure problems.

David Levinson, a transportation analyst and professor at the University of Minnesota, brings up a number of other concerns about this plan. PPPs are complicated multi-decade financial arrangements, and not all states and localities are necessarily well-equipped to manage these deals in the public interest.

Meanwhile, these tax credits would do nothing to attract investors without any federal tax liability, such as pension funds, endowments, and international sovereign wealth funds. “That’s potentially important,” Levinson says. “If you look at the major investors in existing quasi-privatized US tolls roads, they tend to be international players and pensions funds.”

Finally, the Trump campaign’s claim that its tax breaks will pay for themselves by creating new tax-paying jobs looks dubious, Levinson notes. Right now, unemployment is extremely low. Anyone who works on these new privately financed infrastructure projects is likely to be employed already — this would just be shifting jobs around, not creating new jobs. (Levinson did add, though, that it might be worth trying out Trump’s tax credit scheme on a small scale — just to see how it worked.)

My key points

  1. Opportunity Costs.  The assumption about tax recovery assumes the capital and labor would otherwise be idle. This might be true in a recession. In current conditions, they are working on infrastructure instead of doing something else, so only a fraction of the claimed tax
  2. Non Taxable Revenue Sources. Tax credit scheme cuts out many potential sources that don’t pay taxes, like Pensions and Overseas money. In fact, much of the current private investment in infrastructure is supported by international money, which is great (we get the stuff, they get pieces of paper).
  3. Infrastructure tends to be a money loser in this environment. I am not convinced there are that many good investments in infrastructure, not which will pay the rate of return needed to justify the risk. That’s what I think can be discovered with a small pilot. Does anyone actually bite? The problem is not that private infrastructure could not be profitable if everything else where unsubsidized, it’s that everything else is subsidized, which makes it harder for selected projects to be profitable.
  4. Straight-up privatization (long term leases) is simpler. We could just sell or lease off much infrastructure (airports, ports, water utilities, freeway networks) and get a huge up-front lumpsum payment along with continuing revenue. Regulated utilities are a fine model used in most other countries for these services.

Yes Virginia, There is a Successful Macy’s

When recently in New York I visited the pantheon of American retailing, the Herald Square R. H. Macy’s department store, the flagship of the chain and the largest department store in the world, almost the size of the original Mall of America.

img_2942I first went as a teenager, and was impressed by its massiveness and its wooden escalators.

I have been back a couple of times since then. It is the local Macy’s store on steroids of course. There is more of everything, packed more densely. But it has many of the same departments, as well as many more.

It was mobbed. Granted it is November, but it is still more than a week before Thanksgiving.

Its mobbed-ness cannot be said of other Macy’s flagships, which are under threat of closure. Most notable for me is the Minneapolis store, which may follow the sad fate of its eastern neighbor in St. Paul. I have only been to the downtown Minneapolis store a few times, we bought a suitcase there the second day the Hiawatha LRT was open (June 27, 2004 I recall), because the ride was free and we wanted to try it out, and we wound up downtown and needed a suitcase. In short that is the problem, we seldom go downtown. It’s not just the parking, but it certainly is easier to get in and out of Rosedale from where we live, or MoA. And downtown, especially along Nicollet is deadish, especially compared to any street in Manhattan, even with its skyscrapers.

Obviously Manhattan has about 10 times as many workers and residents as downtown Minneapolis. But it also has 10 times as much retail it seems. So that should be in balance.

There were once many more department stores in downtown Minneapolis (so legend speaks), but times have changed. Given that downtown Minneapolis is over the long-term more or less stable in employment, and rapidly gaining residents, one would think there should be more not less demand for retail. So Macy’s in Minneapolis is doing something wrong. This is not unique, lots of other cities have the same problem. One certainly could throw some blame at online retail, and far more at suburbanization of retailing and especially the rise of big boxes, which occurred everywhere, though perhaps less in New York than elsewhere. Yet Macy’s in New York seems to thrive.

I believe Macy’s made a huge mistake going with national branding rather than retaining local brands (Dayton’s in Minneapolis, Marshall Field in Chicago, and so on). Like sports teams and newspapers, people want a local touch, not a national chain, for their flagship stores and their shopping malls. They used to talk pride in their home town department store, cherishing the small differences between their local store and the chain. Now it is not the Best Dayton’s in America, it is a second-tier Macy’s store. Minnesotans feel far less loyalty and reason to spend money.

I will also add that Macy’s situation belies the New York Superiortarian Anthem “New York, New York” If you can make it here (there) you can’t necessarily make it anywhere else.


Random thought. I also went to Bloomingdale’s as a teenager. I remember buying a book “Unbuilt America“, damaged and on remainder there while I was in high school. Department stores used to be much cooler somehow.

Developing a Comprehensive U.S. Transit Accessibility Database

Recently published:


This paper discusses the development of a national public transit job accessibility evaluation framework, focusing on lessons learned, data source evaluation and selection, calculation methodology, and examples of accessibility evaluation results. The accessibility evaluation framework described here builds on methods developed in earlier projects, extended for use on a national scale and at the Census block level. Application on a national scale involves assembling and processing a comprehensive national database of public transit network topology and travel times. This database incorporates the computational advancement of calculating accessibility continuously for every minute within a departure time window of interest. This increases computational complexity, but provides a very robust representation of the interaction between transit service frequency and accessibility at multiple departure times.

Buckshot: Our Future Infrastructure w/ Professor David Levinson

I was on Buck Sexton’s radio show for 15 minutes or so yesterday. We talked about infrastructure. The SoundCloud link is here.

Elements of Access: Meshedness

Urban planners and engineers have long been interested in measuring street connectivity and typically do so with relatively simple measures such as the link to node ratio (called the Beta Index in the Transport Geography field), which divides the total number of links (i.e. road segments between intersections) by the total number of nodes (i.e. intersections including dead ends).  In the above image, the connected network has link to node ratio of 1.6 while the dendritic network’s link to node ratio is 1.0 (a link to node ration of 1.4 is typically considered a well-connected street network).


The connected node ratio divides the number of connected nodes (i.e. nodes that are not dead ends) by total number of nodes (Handy, Paterson, & Butler, 2003).  The networks above have a connected node ratio of 1.0 and 0.6, respectively.  The underlying intent is distinguish between well-connected or gridded street networks and dendritic, treelike networks  – as highlighted in the figure above – in researching relevant issues such as travel behavior, road safety, VMT, and public health outcomes.

Elements of Access: Transport Planning for Engineers, Transport Engineering for Planners. By David M. Levinson, Wes Marshall, Kay Axhausen.
Elements of Access: Transport Planning for Engineers, Transport Engineering for Planners. By David M. Levinson, Wes Marshall, Kay Axhausen.

Topology takes a slightly different approach to understanding this issue.  The Meshedness Coefficient, for instance, measures connectivity by looking at the number of cycles in the network with respect to the maximum number of cycles (a cycle is a closed path that begins and ends at the same node with no fewer than three links).  A Meshedness Coefficient of 0 represents full tree structure (i.e. no cycles), and 1 represents complete connectivity (i.e. every node is directly connected to every other node, which is not feasible in a large surface transportation network) (Buhl et al., 2006).  In non-planar networks, this measure is also known in Transport Geography as the Alpha Index.  The Alpha for the connected network above is 0.4 and for the dendritic network, it is just 0.03.  For large networks, Beta and Alpha are highly correlated.


Xie and Levinson (2007) developed another useful metric called Treeness.  Instead of counting the number of cycles, Treeness is instead calculated by dividing the length of street segments not within a cycle by the total length of street segments.  The Treeness measure also provides a value between 0 and 1, but in this case, the higher number represents a more treelike or dendritic network (Xie & Levinson, 2007).

DOWNED TREE TRAPS MORE THAN 50 HOUSEHOLDS  Image from The Oregonian (Florip, 2010)
Image from The Oregonian (Florip, 2010)

Networks with good overall connectivity are called integrated networks.  Networks with low connectivity are called fractured networks (although fractured networks can still be comprised of connected components).  Again, these measures relate to issues of resilience.  When a single node failure can significantly erode network functionality, the system is fragile.  The image below shows a fallen tree in Lake Oswego, OR that cut off more than 50 families from the outside world (or more specifically, the cars of more than 50 households were trapped) (Florip, 2010).  If only that network had a little less Treeness.