All posts by David Levinson

Nature is Rebounding: Land- and Ocean-sparing through Concentrating Human Activities

I subscribe and listen to the Seminars About Long Term Thinking Podcast by the Long Now Foundation.

So I was surprised and pleased to find the most recent episode featured Jesse Ausubel, (previously mentioned on this blog, and who has a lobster named for him) discussing fascinating trends in dematerialization, and how the environment is recovering (mostly) even without concerted public policy.

The blurb says:

In the field of environmental progress the conflict between anecdote and statistics is so flagrant that most public understanding on the subject is upside down. We worry about the wrong things, fail to worry about the right things, and fail to acknowledge and expand the things that are going well.

For decades at Rockefeller University Jesse Ausubel has assembled global data and trends showing that humanity may be entering an exceptionally Green century. The most important trend is “land-sparing”—freeing up ever more land for nature thanks to agricultural efficiency and urbanization. Ausubel notes that we are now probably at “peak farmland“ (so long as we don’t pursue the folly of biofuels). Forests are coming back everywhere in the temperate zones and in many tropical areas, helped by replacing wild logging with tree plantations. Human population is leveling rapidly and we are now probably at “peak children.” Our energy sources continue to “decarbonize,” and a long-term “dematerialization” trend is reducing the physical load of civilization’s metabolism.

In the ocean, however, market hunting for fish remains highly destructive, even though aquaculture and mariculture are taking off some of the pressure. In this area, as in the others, rigorous science and inventive technology are leading the way to the mutual flourishing of humanity and nature.


At my conference on the future of methane in Fort Worth last November, I found out about a documentary called Switch, about the future of energy in the US. I finally got around to getting and watching.  It is highly recommended for those interested in the subject.

It is largely non-partisan (though the protagonist is a geologist) and covers all of the important energy types in a serious and engaging way, with a bit more meat and less chrome than the typical PBS science show. You have to request a copy from the website, it is not available on iTunes or Netflix yet.

I don’t agree with everything.  I think it overestimates how long the switchover to renewables + nuclear being the largest share of energy  will take. It does this because it underestimates future technology advances — looking more linearly and less exponentially/logistically in terms of technology development and deployment. Similarly it underestimates the ability of larger inter-connected networks to mitigate reliability/availability problems from solar and wind, and advances in storage of various types. I don’t think this change will be overnight, but I hope it will be sooner than the 50 years the film estimates. The cost curves on solar and wind are getting very competitive, and the more interest they have, the more investment they will get.

Seminar: Gravity Models, Information Flows, and Inefficiency of Early Railroad Networks

History of Science, Technology, and Medicine – Spring Colloquium 2015

Friday January 30 at 3:35 p.m. in room 131 of the Tate Laboratory of Physics on the Minneapolis campus of the University of Minnesota (refreshments at 3:15 p.m. in Room 216).

“Gravity Models, Information Flows, and Inefficiency of Early Railroad Networks”

Andrew Odlyzko, School of Mathematics, University of Minnesota

ABSTRACT: Gravity models of spatial interaction, which provide quantitative estimates of the decline in intensity of economic and social interactions with distance, are now ubiquitous in urban and transportation planning, international trade, and many other areas. They were discovered through analysis of a unique large data set by a Belgian engineer in 1846, at the height of the British Railway Mania. They contradicted deeply embedded beliefs about the nature of demand for railway service, and had they been properly applied, they could have lessened the investment losses of that bubble. A study of the information flows in Britain, primarily in the newspaper press, provides an instructive picture of slow diffusion of significant factual information, the distortions it suffered, and the wrong conclusions that were drawn from the experience in the end.

Light Years of Travel

One light year is approximately 10 trillion km  (actually 9.46 trillion km) (or 6 trillion miles (actually 5.88 trillion miles)

Total vehicle travel in the US  is approximately  or 3 trillion miles (2.968 trillion miles in 2012) . So in the US, vehicles travel 1/2 of a light year each year.

The nearest star (which is not the Sun) is Proxima Centauri, which is 4.2421 light years away.  If instead of driving around hither and thither, mostly in circles, we established a giant relay, we could cover that distance in 8.4 years of American driving .

Now of course, it takes a year for light to travel a light year, and cars don’t yet travel at or near (much less faster than) the speed of light. So at an average speed of 100 km an hour, it would take much longer (about 11.4 billion years) for one car to actually drive from here to Proxima Centauri.

This example would require somehow prepositioning all the US cars along the way (say at 100000 km intervals), and the letting them do their annual driving for 8 years (i.e. each would drive 100000 km). This then requires us to construct a space highway, say the Hoober-Bloob Highway, or perhaps a very short section of the Vogon Intergalactic Highway.

At any rate, this gives you a sense of the scale of space .. it is so large, all US traffic only goes 1/8 of the way to Proxima Centauri; or it is so small, it only takes US traffic 8 years to go to Proxima Centauri.

Is reducing negative externalities a positive externality? (Or on Biking and Vikings)

A comment on a recent post on suggested that bicycling has positive externalities, and thus we (society/government) should subsidize it. The same argument applies to many things for which government subsidy is requested, including large stadiums for professional sports teams.


Zeroth, bicycles are not public goods. Bicycles are both excludable (with locks when not ridden, by the rider when ridden) and rivalrous (my riding prevents you from riding it), and are thus private goods. Bicycle lanes may be public goods, since we don’t generally have effective mechanisms to exclude people who do not pay from using bike lanes (unless we register and tax bicycles, and enforce this, which is difficult with present technologies given their long and open nature) nor are bike lanes generally congested. The facility such lanes parallel may in fact be congested (rivalrous), and thus the demand for the space they use may be rivalrous – this usually varies by time of day.

First, bicycling has lots of benefits for bicyclists, otherwise they would not choose to ride their bike. So there are private benefits. (Similarly, professional sports teams are already profitable.)

Second,  based both on theory and evidence, the provision of bicycle infrastructure increases bicycle use and thus private benefits. (Similarly, a new stadium will increase attendance or ticket prices at a sports event.)

Third, what a bicyclist would do it they are not bicycling is not obvious. Perhaps they would walk, or ride transit. Perhaps the trip would not be made (most trips are not work trips, most bike trips are at least in part recreational). Does more bicycling actually result in fewer people using other modes? The classic (logit) mode choice model implies that it does, though the shares are not 1:1, so we cannot assume that each bicyclist would otherwise drive. The IIA assumptions (Independence from Irrelevant Alternatives) implies that if bicycling were somehow to disappear, bicyclists would use other modes in the proportion they have today, so maybe 60% of bike trips would switch to auto driver. More sophisticated models can may be able to answer this question more accurately.

To the extent that fewer people drive or ride transit, at the margins additional bicycling reduces the negative externalities of those other modes (congestion, pollution, noise, etc.). (Similarly, what a fan would do were they not attending and spectating at a football game is not obvious. There would be at least some other leisure activity consumption.) I don’t believe that is sufficient to claim them as positive externalities for reasons discussed below.

Externalities: Definitions

[For a primer, See: Khan Academy on Positive Externalities.]

Economics makes a distinction between Technical vs. Pecuniary externalities. Pecuniary is the SAT word for relating to money, and the term Pecuniary Externality covers cases where Alice’s presence in the market raises costs for Bob.  If we have an upward sloping cost curve, as we move along  the demand curve to the right in quantity, the equilibrium price rises. Alice’s presence in a market (where variable costs are rising more than average fixed costs are falling) raises prices for other consumers (all the Bobs).

The Vikings’ presence in Downtown East drives up land rents for others. This Pecuniary externality is not what we are concerned about. Instead we are concerned about Technical Externalities, where my presence in the market creates an impact on someone who is not in the market.

For instance the noise and congestion from the Vikings game lowers the property value of nearby residences. A bike trail in my back yard may lower my property value. Taking lanes from cars for a bike lane may increase congestion in the remaining car lanes.

Just about any action has positive and negative effects on others who are not party to the transaction. In the absence of regulation, parties try to internalize the benefits and  externalize the costs. The positive externality argument has been used to justify subsidizing many different goods and services.

What is our baseline? 

The Polluter Pays Principle says the baseline is no negative externality. When there is a negative externality, the polluter must pay a tax (a Pigouvian tax) equal to the cost of the externality to society. If reducing the externality can be done for less cost than the tax, the polluter will of course do that.

If instead society is subsidizing someone to not pollute, for instance by giving them a property right in the pollution they are already creating (such as a cap-and-trade program, or congestion credits), we are reducing the negative externality, but systematically distorting incentives all around. We are changing the baseline from no pollution to the existing amount of pollution. We are in a sense giving a property right to existing polluters to continue polluting (congesting) the amount they want , and the victim of the pollution is  required to pay them to stop. While Coasian bargaining suggests we would wind up with the same efficient equilibrium (if we can ignore transaction costs), this will provide a poor set of incentives, especially if the polluter can game this system to increase their initial endowment by maximizing initial pollution in a way that is easily reduced for the polluter, but not obviously so to the pollutee.

Currently cars do not pay for the pollution, congestion, noise, or unsafety they produce.

Is there a dual problem?

If the Polluter Pays Principle is our operating logic for negative externalities, what is the mirror or dual for positive externalities: The Positive Spillover-er is Subsidized Principle? The person or organization who creates positive spillovers (positive externalities) for society which they cannot themselves internalize, should be subsidized.

Assuming we could actually determine this (i.e. assuming perfect information), and that transaction costs were very small (i.e. we could implement it) this might make some sense. Those two assumptions are at odds with how reality often works. Nevertheless, it is certainly used as a rationale for public subsidies for certain things like parks, schools, transit, bike lanes, NFL stadiums, even roads.

Unlike negative externalities which we clearly do not want, no one (or not everyone) necessarily asked you to create a positive externality.

  • If we subsidize to induce the positive externality – then we are asking.
  • If we subsidize to reward the positive externality – then we are thanking, and implying an implicit incentive for future Spillover-ers. But these  are different cases (before vs. after) with different results.
  • If it doesn’t occur to society to incentivize you or thank you, but instead if you are demanding a subsidy  because of the benefits you provide that the rest of us don’t (at first, or ever) see, that is another case still.

There are few cases where everyone benefits, and fewer where everyone benefits the same amount. A Vikings stadium benefits place A and businesses there in large part at the expense of place B and businesses there. In short, most of the benefits are transfers, and while there may be a net social increase, there are winners and losers and the winners do not actually compensate the losers.

Internalize it

While the Negative Externalizer has no incentive to capture their externality, why can’t the  Positive Externalizer capture the positive externalities? The Vikings, for instance, could have moved to a location [Arden Hills] where they put all the parking, hotels, shops, and restaurants on site, and the  only spillover would be name recognition and municipal pride in a team so great they consistently bring home national championships regularly for the metro area they purport to represent. By moving to (staying at) a smaller site in the city, they cannot capture all of the excess spending by stadium-goers, which instead spills-over to neighboring blocks in downtown, enriching nearby landowners and their tenants, and indirectly increasing the tax base (so it is said).

So what is the difference between a request for more subsidy for  bike lanes vs. more subsidy for a professional sports team?

Roads, and wide linear infrastructure in particular, are notoriously hard to privately build without government consent or granting of eminent domain powers. Stadiums as point facilities are much simpler. A new network of bike lanes divorced from existing networks of rail and road infrastructure in a built-up area is impractical. In contrast, a new network of bike lanes in a private master planned community built upon a green field is readily accomplished.

In the absence of master planned communities replacing administratively obsolete cities cursed with an excessive division of property, a practical solution in the messy city needs to be identified.

Is the good undersupplied when paid for by direct beneficiaries?

Would we get fewer or smaller football stadiums when the Vikings pay for it instead of the public, and would the public thus lose benefits. One could hope. Certainly the Vikings have gamed the system to get a huge reward for little investment.

Many goods have positive externalities over some range of quantity, but are not necessarily undersupplied once fully deployed. Network externalities are an example. My use of the cell phone network makes it more valuable for others. My taking the bus makes bus transportation more valuable for others (See Mohring Effect). Similarly, my taking flights out of Minnesota makes air travel more convenient for others in the long run, as more flights will be supplied through Minneapolis, and hubs provide greater connectivity.

That doesn’t necessarily mean that the public should subsidize my cell phone, bus trip, or  flight even though I provide this externality I cannot capture myself. The private benefits are large, and after a certain point, public subsidy would not actually induce more consumption since the consumption is at the maximum level economically feasible. This is certainly close to true for cell phones  where networks were subsidized privately rather than publicly. (There has obviously been significant public subsidy in the aviation sector, though today it is mostly privately funded). It is widely debated for public transport.

How you do the subsidy matters

We can subsidy the consumer directly, but giving them cash or tax credit. Alternatively, we can subsidize supply, lowering the cost and thereby inducing more demand. These have very different effects.


Do Two Wrongs Make A Right? Do Two Minuses Make  A Plus?

So back to the original question: Is a reduction of negative externalities from another mode (a benefit to society as a whole, if not necessarily to each member) a positive externality of biking?

Without immediately answering, we can say that, if a reduction of a negative by subsidy is a positive, it is a “second-best” solution. The first best solution is to reduce externalities directly by taxing them so as not to distort incentives and to discourage over-consumption directly. However, as wikipedia says: “In theory, at least, it may be better to let two market imperfections cancel each other out rather than making an effort to fix either one.” The imperfections in this case being the negative externalities that are not internalized, and the subsidy for a mode that might reduce those externalities.

Classic economic examples imply that it is a positive externality. One random website writes: “If you walk to work, it will reduce congestion and pollution, benefiting everyone else in the city.” which is about as analogous as you can get. Yet something rings wrong.

The underlying logic is that “negative” implies the minus sign, and if we reduce a negative we are adding (two minuses make a plus in math even if two wrongs don’t make a right). From a simple welfare economics perspective, you will get the same Net Present Value either by more biking because you subsidized biking or you taxed non-biking.

NB: You will not necessarily get the same Benefit/Cost ratio, since negative externalities are a cost and positive externalities are a benefit, and the subsidy is a cost (to the government), and a benefit (to the traveler) and tax revenue a benefit (to the government), and a cost (to the traveler).

However, as we noted above, everything has tons of indirect effects. Lifecycle analysis often look at these. When thinking about externalities, we need to distinguish between direct and indirect effects. To illustrate, when driving a car, tailpipe emissions are a direct externality. You may have purchased the car in a market transaction. The car was manufactured in a factory. The factory also had emissions. Are the factory emissions a negative externality of driving a car? If they are, how about the emissions of the steel factory? How about the emissions of the worker who drove to the steel factory? How about the emissions of the food truck that supplied the steel worker’s lunch? How about the emissions of the clothing factory that produced the shirt which was on the back of the driver of the food truck?  If we accept the auto factory, we have no basis not to accept  everything else in society, since the entire economy is a connected network. The usual rule in economic analysis is that we look at direct effects, not indirect effects mediated by market transactions. In this way we can focus on real effects and avoid double counting.

Thus reducing negative externalities of driving because you walk to work may be a wonderful thing, but contra our random internet website it is not a positive externality of walking any more than the negative externality of the emissions of the factory worker driving to the shoe factory which supplied you with shoes. And any positive externalities of bikes should be associated with those direct effects of bicycling, not the indirect effects of other avoided things, nor should they be offset by the negative externalities of bicycle manufacturing and distribution.

I am not sure I can identify any direct positive externalities of bikes that are not mediated. Perhaps people like to look at bicyclists, who provide a positive aesthetic externality. (Actually, people do like to look at unicyclists, which is why circuses can charge money). Maybe bicyclists are healthier, and reduce claims on publicly subsidized health care. (A. This is not a direct positive externality since it is mediated by the health insurance market, and B. I am not sure this health claim is actually true, especially after considering safety and injury and deeply breathing in the toxic emissions of cars – health effects are context dependent, and what would be true if everyone biked is not true if most do not. See e.g.: Hankey et al.).  Bikes are good, but they are mostly good because they are good for bicyclists. Like other transportation modes, they should be funded primarily by user fees. Only if that is not feasible should other sources of funding be considered. [Recognizing of course, that since other modes are not fully funded by user fees, we are in the world of the second-best, and it would be unfair and inefficient for only bicycle facilities to be fully funded by user fees.]

So then the question is whether these indirect positive effects are worthy of subsidy.

Political Economy

One of the problems is when the private beneficiaries are clearly identified (and organized), but the public beneficiaries are diffuse. This is a classic political economy problem, and explains why many special interests get tax breaks or subsidies. So the beneficiaries have an incentive to overstate the social benefit, especially when there is not a clear neutral arbiter of facts.

The net result may be overinvestment in such facilities on the grounds of positive spillovers. This is obvious in the example of the Vikings stadium, which is far more than actually needed to keep the owners financially compensated for staying in Minnesota. While bicycle advocates will scoff at this premise in the present context, there are many examples of overbuilding in the history of transportation. Just look at paved roads for cars and trucks, which at one time were in the same position as bike-only facilities today, and were clearly under-supplied. Today I suspect most bicycle advocates would assert such roads are over-supplied.

A second issue is the deadweight loss (the social benefit that is missed because of under-consumption) might be very small compared to the private benefit. This depends on the shapes of the curves and the magnitudes of the private and public benefits, both of which are unknown in practice.

Alternatives to Public Subsidy 

Public subsidy is not the only means for groups to obtain what they want when it is infeasible to do so privately. The website produces many social benefits, but is paid for by a relatively few (compared to readers) members. The readers must benefit, as they read. They probably benefit more than the private time and effort involved in reading. Yet they mostly don’t pay, and instead free ride (free read) on the efforts of others, and is produced anyway. Life is a free-ride on the efforts of our ancestors, who brought us fire, steam power, electricity, and the internet, among millions of other innovations.

Many philanthropic activities so occur.

Why isn’t transportation itself philanthropically funded? While there have been discussions of philanthropically subsidized public transport, I am not aware of this actually being implemented at scale. One may argue the scale is too large. While that may be true for large public works, surely that is not true for bike lanes, which are relatively inexpensive to construct. (Bikeshare systems are often funded with a mix of  philanthropy and user fees).

More likely it is because government already owns the roads, and thus the space out of which most bike lanes would be carved. Government is currently charged with building and maintaining roads, so that is the status quo. And if you can convince someone else to pay, why do it yourself.




Transit shelters and other amenities affect perceived wait times

CTS Catalyst reports  Transit shelters and other amenities affect perceived wait times on our soon-to-be-published report.


The overall results: “Perceived and actual wait times are clearly related, but the relationship is variable,” Guthrie says. “The waiting environment can change perceptions.”

Nearly 85 percent of those surveyed waited 10 minutes or less. Even with waits under a minute, however, people tended to perceive at least a minute or two, and they tended to estimate in round numbers (5, 10, 15 minutes). “This creates an initial ‘penalty’ of overestimates,” he says.

Researchers also found several variables to have statistically significant impacts. The presence of a shelter—even a simple one—made waits seem shorter, especially for waits less than 10 minutes. “The biggest difference in perception was between any shelter and none at all,” he says. The presence of a NexTrip real-time information sign also shortened perceived waits.

Posted schedules produced a “really interesting pattern,” Guthrie says. For shorter waits, schedules caused people to overestimate wait time, but after about 10 minutes, people began to underestimate it. “It’s possible that for short waits, people compare the clock and the schedule and get impatient, but for longer waits, they are reassured to know the bus or train is coming,” Guthrie says. “This implies that posting schedules is more important for routes with less frequent service.”

Gender alone was not significant, but there was a stark difference for women in less safe environments. “Most sites in the study were rated as safe, but at those that were not, there’s potential to improve the experience for riders and potential riders,” he says.5 Minute Wait

“With several major initiatives currently under way to expand the number of shelters at bus stops and to improve the quality of transit schedule information across our entire network, the timing of this project could not be better,” says Marilyn Porter, director of engineering and facilities for Metro Transit. “This study provides important insight that is directly applicable to the work that we are doing to ensure that our customers have the best possible experience using transit service in the Twin Cities.”

The model developed in the project includes many other variables such as household income, trip purpose, and the presence of benches and route maps. “Users of the model will be able to choose criteria and predict the impacts of hypothetical feature mixes,” Guthrie says.

A final report is planned for publication in March. Humphrey School associate professor Yingling Fan was the study’s principal investigator; David Levinson, RP Braun/CTS Chair in the Department of Civil, Environmental, and Geo- Engineering, was co-investigator.

Autonomous vehicles: The legal and policy road ahead

CTS Catalyst summarizes some of the discussion from the recent Conference Autonomous vehicles: The legal and policy road ahead


… David Levinson hypothesized some possible directions:

  • Autonomous vehicles enable more car sharing. Instead of the sunk cost of car ownership, people pay the marginal cost per trip—and thus make fewer trips.

  • Shared cars can be right-sized for any given trip, so fewer large cars are needed. Increased safety reassures people about driving smaller cars.

  • Smaller cars travel closely together on narrower lanes, so capacity increases.

  • As networks get faster, people choose to travel farther. Cities decentralize and more megacities and “placeless places” develop.

  • At the same time, inner cities get denser, as less space is needed for parking and garages.

  • With lower labor costs, transit becomes more cost-effective.

  • Driverless trucks lower delivery costs. Combined with drones, robotics, and online shopping, retail shopping declines.