Why we engage in ‘security theater’

This post originally appeared in  Symposium Magazine.

The politics of security are difficult. If you are in favor of security, you must be in favor of more spending on security, or on anything that will “keep us safe.” If politicians or bureaucrats oppose a proposed security measure and something happens, they will be blamed. Security ratchets up quickly. Ratcheting down can only really be by attrition.

Although the increase in security spending post–9/11 exceeds $1 trillion over ten years, the Department of Homeland Security has systematically failed to study the costs and benefits of proposed security measures, most notably radiation scanners. Other Western countries have had similar spending increases.

The problem is not just the total, which is large and might be spent in other sectors. The problem is the allocation within the security sphere. If you have $1 trillion to spend, what is the best way to do that to maximize security? The idea of opportunity cost rises again and again, and is never properly dealt with.

Rather than assessing both the probability of an outcome and its cost if it occurs, the agency has dealt with risk qualitatively, imagining worst-case scenarios and engaging in what the legal scholar Cass Sunstein calls “probability neglect.” The artifice of travelers removing shoes and belts, unpacking suitcases, pouring out liquids, while agents frisk grandmas and children at security checkpoints has been called “security theater” because it aims to give travelers the impression that something is being done to improve security, when there is no evidence any of this has made a difference.

The vector of attack from 9/11, hijacking planes and crashing them into buildings, has been unrepeated not because TSA has ensured safety. Instead, secured cockpit doors and the willingness of passengers to take action in the event of a threat on board has prevented this from occurring. Other attack vectors that remain are receiving less protection because so much money is spent in response to the previous attack. While the TSA monitors security, the lines leading up to security are managed by the airport or airlines themselves, and the latter give priority to certain passengers (e.g., first class), even though they pay exactly the same amount of security tax as everyone else.

This has some analogy with HOT lanes; the key difference is that the car in the HOT lane pays more money to go faster, while in this case, first class passengers only paid more money for a better seat on the plane. Security is not just an issue on airplanes. Inter-city buses and high-speed rail sometimes have security measures, though rarely as stringent. Intra-city public transit generally does not, because it would quickly become unworkable. As a consequence, these other modes are much more vulnerable, as shown in the Madrid attacks (2004), London attacks (2005), or the Tokyo Sarin gas attack (1995).

Security is the enemy of efficiency. Just as with safety, we want perfect security. That goal is unattainable, and security providers should rationally trade off between value of time and value of life.

Excerpted and adapted from The Transportation Experience: Second Edition, by William Garrison and David Levinson (Oxford University Press, 2014)

Why time is money on the road

This post originally appeared in Symposium Magazine.

In my post yesterday, I described how my team and I set up an experiment to assess the conversion of HOV lanes to HOT lanes – basically, a move that asks drivers to decide how much they want to pay to drive at peak times. We recruited subjects who live along the I-394 corridor in Minneapolis and had a daily route of at least 20 minutes so that the alternative commutes would make sense, assuming they were shorter.

The other conditions were that the I-394 route had to be a plausible alternative for them; the drivers had to commute on a regular basis; they had to work near downtown Minneapolis, and they had to drive a car alone. We obtained permission to install a GPS unit in their vehicle, and they had to follow our instructions about which route to take during which weeks.

We then gave them three sets of alternative routes. One was the HOT Lane on I-394; the second was the general-purpose (un-tolled) lanes; and the third was a parallel arterial, which depended on where they lived, and which would be most feasible. We asked them questions every week about what they were doing and gave them two weeks of free choice to establish their baseline preference. After that, we divided the next six weeks into three two-week periods in which they drove on each of the three alternatives. For the final two weeks, travelers were again free to pick a route, allowing us to see whether their behavior changed after the experience.

The study’s aim was to ascertain what we call the “Value of Travel Time Reliability.” This is how much of a premium drivers are willing to pay to have a low variation in their travel time. The advantage of the HOT Lanes is that they don’t just offer a lower travel time, but more importantly, they offer a much more reliable or predictable travel time. Indeed, they vary much less than the congested general-purpose lanes.

Depending on how we measure variability, we got estimates of value of time ranging from $9 to $20 an hour – in other words, the premium a driver is willing to pay for speed. As for the “price” of reliability, we got values ranging from $3.80 to $18.23 an hour. Other people have tried to estimate a reliability ratio using a variety of methods, and the numbers vary widely. But they tend to be on the order of 1.0, indicating that a driver considers avoiding a minute of unpredictability is as valuable as avoiding a minute of expected travel time.

Why HOV lanes often don’t work

This post originally appeared in  Symposium Magazine.

In my article, I outlined a series of experiments that I and my team have run over the years that explain how people make decisions – some rational, some not – about their daily transportation habits. Today, I’m writing about one of the most popular traffic innovations in congested areas: High Occupancy Vehicle (HOV) and High Occupancy Toll (HOT) lanes.

HOV lanes are designed to expedite buses and cars that carry multiple passengers (and oddly, motorcycles). But there is often insufficient demand by drivers to fully utilize HOV lanes, while general-purpose lanes remain congested during the peak periods. This is inefficient.

If we simply opened up all lanes to all vehicles, we would lose the time saved for HOV vehicles and bus passengers. It would also make it difficult to restore HOV status to the lanes in the future — and people are more averse to losses than they are open to seeking gains. So the city of Minneapolis implemented a solution to transform the HOV lanes to HOT lanes on the I-394 corridor.

HOT lanes allow vehicles equipped with transponders to use what were the HOV lanes for a price. This toll varies with traffic conditions, but it aims to ensure that the HOT lanes maintain lower travel times than the general-purpose lanes. Travelers in a hurry might be willing to pay a premium to guarantee they can avoid congestion. Other travelers won’t.

Joined by Kathleen Harder and graduate students Shanjiang Zhu and Carlos Carrion (now a post-doc at the Singapore MIT Alliance for Research and Technology), I ran experiments to see how effective this switch really was. The aim was to use each commuter’s actual origin and destination but making sure they experienced alternative routes rather than staying bound to their usual route. We found out that there is a way to “price” an efficient commute – and I’ll write about that tomorrow.

Understanding trade-offs and public trust

This post originally appeared in  Symposium Magazine.

I’ll start my posts this week by making some general points about how people view trade-offs when it comes to public dollars — not just for transportation, but infrastructure, more generally. This topic was the subject of recent interviews run by Chris Riback on “Conversations with Thinkers,” which you can view here.

First, we need to understand that cities and municipalities make trade-offs in different ways. We all have different priorities. Minneapolis, where I live, has just built a new baseball stadium for the Twins, a new college football stadium for the Golden Gophers, and is about to replace the Metrodome for the Vikings, for example. This shows how important sports are to us, and that there is political will to achieve this (even to the point of violating local democratic checks). At the same time, we are underfunding many other things, including maintenance of our crumbling roads.

So if the government directs money to priorities that are different from what we think they should be, there is an erosion of trust in government. Citizens think the government has priorities that are different from what they think they should be, and they think government doesn’t spend their money well. We need a mutuality of trust for citizens to understand when it’s time to spend money on some shiny new object that they want, and when it’s time to invest, say, in resurfacing roads or rehabilitating bridges.

Of course, this leads to the fundamental issue of taxes. People don’t like to pay them, even if we save much more money up front by investing in infrastructure than by paying for repairs later on. Nationally, we have lost something like 20% of the numbers of transportation construction workers since the start of the recession as federal and state dollars for investments have been pared back. So we have to address the question of what we can afford given these constraints on finances.

Bikes still outsell Cars in US

The End of Traffic and the Future of Access: A Roadmap to the New Transport Landscape. By David M. Levinson and Kevin J. Krizek.
The End of Traffic and the Future of Access: A Roadmap to the New Transport Landscape. By David M. Levinson and Kevin J. Krizek.

A recent meme was going around Twitter noting that in Europe, bikes were outselling cars (NPR report) (Part of the problem is the misleading headline in this Time Magazine rehash of the NPR report). This seemed obvious to me, and I am surprised it was news, since it is true in the US as well. I tweeted to the effect:

  • US Bikes: 18.7 Million NBDA
  • US Car sales 8 million, US light truck ~ 8 million  … WSJ

This was widely retweeted. CelloMom commented:

“Even if are children’s bikes, still at parity.
So where ARE all these bikes, why don’t we see them on the roads?”

Bill responded:

“gathering dust in garages”

The meaning of this statistic is clear. Americans like to buy bikes. Just ask Kevin Krizek, who was rumored at one point to have a quantity of bicycles running into the double digits.

Of course many of the bikes are kids bikes (5.7 million of the 18.7 million are below 20 inches wheel), but even so, 13 million are 20 inch and above wheel size, and 13 million is still much bigger than 8 million cars (and near 16 million light vehicles, note also many light vehicles are not for personal use). Even if we just look at specialty bike shops, which sell at the higher end, that’s nearly 3.1 million bikes per year, which while less than cars, is still a pretty big number.

Yet, the number of trips by bike and certainly miles by bike are much lower than by car and are not poised to overtake in the US. We don’t even see 3.1 million bike commutes daily in the US (ACS reports 865,000), so these are more likely for recreational than utilitarian purposes.

Another reason for this statistic is that bikes don’t last as long as cars (The average US car on the road is 11.4
years; I could not find similar data for bicycles, but am sure it is lower, especially given the higher sales — at 18.7 million bikes per year there would be 1 bicycle for every person in the US every 16.7 years, so the average age would be about 8.4 years IF everyone had a bike and there were no losses, and surely that isn’t true). This again is in large part due to the growing up of kids. Reasoning from anecdote and personal experience, (always a bad idea) our garage has 1 “light truck”, 2 striders, and 2 bikes. By next year there will be at least 3 bikes (and maybe 2 more if the adults get them again).

Fleet Turnover

The private automobile fleet in the US turns over very slowly. This is because as capital, it is not used most of the time, and thus wears out slowly. If instead of 250 million vehicles operating 1 hour a day, suppose we had 125 million vehicles operating 2 hours per day, or 62.5 million vehicles operating 4 hours per day. Our vehicles would be replaced 4 times as often (assuming they wear out with usage), and the average age of vehicles on the road would be under 3 years instead of 11.4 years. I have had 7 (different generations of) cell phones in the past 11.4 years.

We (as both individuals and society) want a newer fleet because newer cars:

  • are more fuel efficient,
  • pollute less (both because they use less energy and because they have better pollution control),
  • are safer,
  • have better user interfaces (we hope) and are more in-sync with changes in information technologies.

We (as a society) want fewer cars because that:

  • requires less parking, and
  • makes driving less likely.

This efficient use of capital (keeping vehicles in motion 24/7) is a hallmark of large, expensive fleets like shipping, airlines, railroads, and to a lesser extent trucks.

Instead of paying a fixed cost of ownership once (independent of use), and a variable cost that includes only fuel and time, the cost of car usage would include paying for the fixed cost of ownership on a per trip basis. This would significantly raise the out-of-pocket cost of driving, and discourage it, but also make driving better. But it might also lower the total cost of transportation, since individuals would no longer have so much capital tied up in vehicles, and would drive more efficient cars, less often. This is independent of, and multiplicative with, any reductions in vehicle use that could arise with increased ride-sharing enabled by logging your planned trips in advance.

We can achieve this with Cloud Commuting, Car sharing where the vehicle comes to you. But even before the halcyon days of driverless cars lift up humanity from the need to be ever alert while traveling, ubiquitous car sharing where the vehicles are omnipresent instead of rare would make this much more feasible even in medium density suburbs.

Car sharing has strong network effects. I am more likely to use car sharing if my neighbors use it, since that makes it more likely there will be a car in front of my house, my workplace, my shop, or wherever, when I want it. A reduction in vehicle access time from 10 minutes to 5 minutes, or 5 minutes to 2 minutes is very significant, especially when most trips are only 20 minutes long. As with any social network, it is not clear in advance which if any will take off. As with many networks, there needs to be a large up-front capital investment. But unlike transit systems, car-sharing is dealing mostly with mobile capital. If the program doesn’t work in place A, cars can be redeployed to place B, or at worst, sold in a used car lot.

So the economics of sharing makes sense, but the sociology of sharing is still unclear. People will share hotel rooms, or bikes, or library books, but not many other goods (or historically cars). How do cars get transformed from an owned good to a rented service? In part this is generational. If you have never owned a car, new habits can be formed. But that type of change is very slow. Early adopters and the carless may be quick to join. Some use their cars often enough, in places remote enough, or customize their cars sufficiently that carsharing will not be advantageous. Where is the threshold? I am on record as a skeptic (even if it is a good idea). The Car2Go model (of which I am a member, but which I have yet to use), which recently invaded Minneapolis, where the cars can be left in on-street parking rather than returned to the base seems progress, maybe they will put in enough capital so there is a car waiting for me on every block. But there are a lot of blocks in Minneapolis (1100 miles, I estimate some 11000 block faces), so moving from 212 cars to some 10000 (as a rough approximation of where it needs to be so I don’t have to walk more than a block to find one) is a 50-fold expansion. While reviews are favorable, finding one of 212 cars on 1100 miles of street is not going to be a dominant mode. (or roughly 4 per square mile in the city, meaning roughly a 1/4 – 1/2 mile walk to get one, which of course varies depending on where in the city you are).

The required 50-fold upscaling to make carsharing approximately block-level is non-trivial (10000 cars at ~ $10000 each is $100,000,000 (still well less than a Vikings Stadium, or on the order of a single Streetcar line!)). But 10000 cars is less than several hundred thousand registered in Minneapolis, and could replace many of them.

Catalyst: The future of transportation

CTS Catalyst summarizes some of my recent clips from The Week.

Levinson predicts improved accessibility, more real-time transportation data, and cleaner cities.

What does the future hold for transportation? University of Minnesota civil engineering professor David Levinson shared his thoughts as part of a panel of transportation experts in a series of videos from The Week magazine. Some highlights from Levinson’s predictions are below.

Urban mobility in 10 years

Over the next 10 years, Levinson predicts the emergence of driverless cars on the road, more car sharing and bike sharing programs, and more real-time information about buses and traffic congestion. “The future is already out there in pieces, but it will be much more systematically deployed in 10 years,” he says.

He also predicts that cities will be cleaner, with more electric cars and lower levels of tailpipe emissions, even for transit vehicles and trucks. “Cities will be more pleasant places to live,” he says.

Mobility versus accessibility

“We’ll need to think about transportation not as providing mobility but providing accessibility,” Levinson says. “It’s not just how fast we move on the network, but about how many things we can reach.” According to Levinson, connecting people to the places they want to go is not only a transportation issue, but also a land-use issue.

The challenge, Levinson says, is that these issues are often governed by different organizations. “Land use is generally locally managed, and transportation is funded at least in part by the federal and state government…They have different objectives,” he says. Improvements in accessibility will require better coordination and alignment of these transportation and land-use objectives.

The role of data

According to Levinson, accurate and reliable transportation data are and will continue to be important because they can provide real-time information to help travelers plan in real-time. And although data are already being used to provide information to drivers, transit users, and flyers, there are still areas where data are incomplete—such as for travel time on urban arterials.

“We’re in process,” Levinson said. “We’re going to be doing a lot better in five years than we are today, and we’re doing a lot better today than we were five years ago. But we’re not there yet in terms of being able to fully exploit the information that’s out there.”

Related Links

Open Access Article: Spatial modeling of bicycle activity at signalized intersections | Institute of Transportation Studies Library

Open Access Article: Spatial modeling of bicycle activity at signalized intersections

Biking at Grand/Halsted/Milwaukee (3 of 4)

This week is Open Access Week. What’s Open Access? Here is a not very brief overview by Peter Suber. UC Berkeley also has an Open Access Initiative to help open up your research and data. 

In the spirit of Open Access Week, here’s an interesting article from an open access journal – The Journal of Transport and Land Use. Go check it out and peruse the articles. No need to depend on your institution’s sibscription because it’s free to the public! (Thanks open access!)

In “Spatial modeling of bicycle activity at signalized intersections“, Jillian Strauss and Luis F Miranda-Moreno look at the built-environment and cycling. 

This paper presents a methodology to investigate the link between bicycle activity and built environment, road and transit network characteristics, and bicycle facilities while also accounting for spatial autocorrelation between intersections. The methodology includes the normalization of manual cyclist counts to average seasonal daily volumes (ASDV), taking into account temporal variations and using hourly, daily, and monthly expansion factors obtained from automatic bicycle count data. To correct for weather conditions, two approaches were used. In the first approach, a relative weather ridership model was generated using the automatic bicycle count and weather data. In the second approach, weather variables were introduced directly into the model. For each approach, the effects of built environment, road and transit characteristics, and bicycle facilities on cyclist volumes were determined. It was found that employment, schools, metro stations, bus stops, parks, land mix, mean income, bicycle facility type (bicycle lanes and cycle tracks), length of bicycle facilities, average street length, and presence of parking entrances were associated with bicycle activity. From these, it was found that the main factors associated with bicycle activity were land-use mix, cycle track presence, and employment density. For instance, intersections with cycle tracks have on average 61 percent more cyclists than intersections without. An increase of 10 percent in land-use mix or employment density would cause an increase of 8 percent or 5.3 percent, respectively, in bicycle flows. The methods and results proposed in this research are helpful for planning bicycle facilities and analyzing cyclist safety. Limitations and future work are discussed at the end of this paper.

The full article can be found here