White Rose • Total Surveillance versus Anonymous Charging: <small>the road pricing dilemma</small>

Tuesday, July 01, 2003

Total Surveillance versus Anonymous Charging: the road pricing dilemma

Brian Micklethwait (London)

(2)

For as long as I can remember I have been an enthusiastic supporter of the principle of road pricing, for much the same reasons that I favour the pricing of any other scarce and desirable product or service. Reduce queueing caused by underpricing. Encourage the construction of better roads, more suited to the desires of drivers, more creatively designed. Pricing will enable road ownership, and that will enable better environmental policies, because owners will then be responsible for environmental impact. Etc.

However, there are two different ways of doing road pricing, both of which have big advantages and big disadvantages.

One. Anonymous Charging. Charge each vehicle to go past certain barriers, physical or electrical. Either the man at the wheel chucks some coins down a shute, or the place has a machine which debits the vehicle as it goes by, by debiting a box on the vehicle which has been filled up with money, gas meter style.

Advantage: Anonymity! The vehicle user is no more spied on than he is when he buys a pair of socks in a shop. If the vehicle user consents to the transaction tracking inherent in the use of a credit card, fair enough. But money remains an option, and money is freedom, because money is anonymous. (I remember once a trader in a street market shouting at me: "You don't ask me where I got the stuff I'm selling, and I won't ask you where you got your money.")

Disadvantage: Cumbersomeness. Every barrier becomes a huge Thames Flood Barrier for cars. Installing machines in cars is complicated and expensive, and what if different cities use different systems? A different box for each system? Until the same system wins a battle of the gauges, it's a nightmare either of delay or of incompatible equipment.

Two: Total Surveillance of vehicle number plates, and bills sent to the owners of vehicles for journeys completed, rather than payment demanded from the driver of the vehicle during the journey.

Advantage: No interruption to the journey. No new kit needed for each vehicle that might need to use the system, only new surveillance kit and a billing mechanism for the system as a whole. (All vehicles already have number plates.) Therefore, such arrangements can be introduced gradually.

Advantage: Much easier in due course to introduce more complicated but more rational pricing systems, based on Total Surveillance of each vehicle journey, involving different prices for different parts of the road system, and for different times of day. Paying "by hand" tends to be too crude. Five quid to pass this big barrier regardless of how much time you spend in the city, or when you do it is the present London regime. If drivers paid cash at the barriers, making this any more complicated would be a nightmare. Total surveillance will allow much more exact pricing systems in the future.

Disadvantage: Total Surveillance! Big Brother spies on you and your movements all the time. Do we trust Big Brother to confine all the knowledge he gathers only to the task of sending you an accurate bill? (Do we even trust him to send out accurate bills, always to the right people and never to the wrong people?)

Disadvantage: stealing a car and driving around in it becomes even more like stealing a credit card. Not only does the victim lose his car. He is all too liable to be billed for the getaway travelling costs of the thief. Anonymous Charging means the driver, whoever he is, pays.

For me, the really serious problem with Anonymous Charging at the point of use is that it is so hard to introduce road pricing schemes based on it a bit at a time. And beware any scheme of improvement that has to be introduced in its entirety before it works properly, I say. Which means that I am inclined, on those grounds, towards the Total Surveillance method rather than Anonymous Charging. Get rough and ready systems going in a few places, such as Central London, and deal with all the unforeseen problems, of traffic flow and of billing and car thieving and fake number plates. Then if the first schemes work okay and the populace is not too enraged and gets the point of it, do more of it, bit by bit, place by place.

But, once the Total Surveillance system is in place, and on a large scale, I can't, off the top of my head, see how The Authorities could possibly be persuaded to switch to the cash-in-buckets alternative, even supposing that they wanted to.

You don't have to attribute malign intentions to those currently administering the first few road pricing schemes in Britain to note that the trend is entirely in the direction of Total Surveillance via number plates of vehicles and the billing of vehicle owners. Power and information gathering is being centralised, because dispersing it will, to start with, be too complicated.

However, once the system is in place, it is simply not credible that the information will only then be used for the billing of road use. The very public itself is liable to clamour for the use, if only on "efficiency" grounds, of all the various spy-cameras, for terrorist tracking, mugging abatement, rape and drug dealing control, dangerous driving dissuasion, etc., etc., etc. In short, the "system", the "authorities", the Powers That Be, would become addicted to Total Surveillance road pricing, and the idea that they would then voluntarily submit to an anonymous, pay cash as you go, system, is simply impossible to believe.

At this point I must acknowledge that this entire piece is a rewrite by me of a piece by David Sucher which I read, and then couldn't find. I only found it again after I'd written the rest of this. He says:

It is simply preposterous to imagine that information gained by Transport for London about movement of specific cars will not (nay "should be" from the perspective of public safety) be shared with the police to track down, say, terrorists. Consider such a possibility when there is a prospective terrorist action. Or in a civil matter, suppose one wanted to defend from a charge of extra-marital dalliance (or wanted to prove it.) You've got a huge data-base of trip movements tied to specific vehicles. The idea that it will not eventually be used for a purpose other than charging for street use beggars credulity.

In response to which this guy, for instance, invents a road pricing based on cash in buckets and anonymity, if that's what you want. He explains how a government that wanted to avoid Total Surveillance could do so. What he doesn't explain is why on earth a government like that of Britain now would go to all that bother, when Total Surveillance can simply be switched on.

All of which means that many civil libertarians will oppose the entire principle of road use pricing.

Personally I think that a permanently chaotic road system is one hell of a price to pay for personal privacy, and for the privilege of the government maybe not being quite so keen to spy on your car journeys. Because don't forget, even if road pricing doesn't catch on, the road spying is here to stay for the time being and for the foreseeable future.

The only way I can ever see a pay as you go system actually replacing Total Surveillance is if, somewhere else in the world, they get such systems going, and thereby, as it were, settle the battle of the gauges battles for us, and give us a working large-scale example of anonymous road use pricing to simply install confidently here, in one great big go. We' all buy our pricing boxes for our cars and stuff cash into them, just as we now buy gas meters or sattelite dishes. Following a civil libertarian revolution lead by White Rose, presumably. Well, you never know, but I don't see it happening right away.

Meanwhile, news of any successful anonymous road use pricing schemes, anywhere on earth, is of immense value to the rest of the world, on road pricing grounds of course, but also on civil liberties grounds.

Comments

Australian toll roads (of which there are maybe 10-15 nationwide) have relatively recently switched from cash in buckets schemes to what you describe as total surveillance systems. (You buy a wireless electronic tag that you attach to your rear vision mirror. This is linked to an electronic account that stores a certain amount of your money, and is updated every now and then by charging your credit card. When you drive through a toll gate the tag makes a beeping sound and the toll is deducted). The typical response of motorists is to find this a huge relief, as it is no longer necessary to slow down and get in a queue to pay the toll. You simply drive along the road and the payment is automatic. Queues to pay the toll at rush hour could in the past delay you five to ten minutes, and they no longer do.

While the number of toll roads remains relatively small, the privacy implications also remain fairly small. The number is only going to increase, however.

Posted by: Michael Jennings on July 1, 2003 11:38 PM

I think Michael's example is closer to what Brian describes as "anonymous charging," at least where the choice is still available to pay using cash. ETC-only schemes are closer to "total surveillance," although usually they do not record number plates (what we in North America call "license plates" or "tags") except if the ETC device fails to register properly or isn't present (i.e. there's a toll violation).

The ETC devices aren't particularly cumbersome using today's technology, although interoperability between toll facilities remains an issue. There is an emerging "EZ-Pass" standard in the toll-dense northeastern corridor states (from Maryland to Maine), but outside the region there's a hodge-podge of incompatible toll devices.

The privacy issues for most users could generally be alleviated by using stored-value systems, but that would require much more sophisticated in-car hardware to be reasonably secure.

Posted by: Chris Lawrence on July 2, 2003 12:06 AM

BTW, interesting tests of quasi-anonymous road pricing will probably be widespread in the U.S. by the end of the decade. The next six-year federal transportation bill is likely to authorize the states to congestion-price single-occupancy vehicles in high-occupancy vehicle facilities ("car pool lanes"), so long as they don't degrade capacity, and to experiment with congestion pricing on other federally-funded highways (see here). And Texas has committed to using toll financing to rapidly expand capacity in urban areas, in addition to an ambitious toll-financed intercity multimodal transportation system called the "Trans Texas Corridor."

I assume all of these schemes will be transponder-based, rather than relying on number plate recognition (which is much harder in the U.S. and Canada, due to the wide variety of number plates issued even within particular states).

Posted by: Chris Lawrence on July 2, 2003 12:14 AM

The option of paying with cash still exists for most of the roads I described, but not for all of them. For instance, the Citylink motorway system through the centre of Melbourne does not accept cash and the electronic tag system is mandatory if you wish to use the roads. New roads opened in the last two or three years have generally dispensed with the cash option entirely. The number plate is not recorded, but the tag is linked to an electronic account which is registered to an actual car, so it might as well be. On the other roads, the number of lanes devoted to cash is declining and at least some of the motorway operators (mostly private companies) have stated that their intention is to eliminate the cash option within a small number of years. The incentive for them to do this is obvious, as it reduces staff costs, plus they get to learn more about the habits of their customers. Helpfully, you get a statement every month or two listing all the toll roads you have used and when, of which the privacy implications are obvious.

We did have more than one incompatible system of electronic devices for a while, but the various operators eventually agreed to sit down and agree on compatible systems with a central clearing house. Of course, the number of different organisations operating toll roads in Australia is much less than in the US, and many fewer than those using the EZ-Pass system in the North East.. Standardisation was thus easier to achieve.

Posted by: Michael Jennings on July 2, 2003 12:24 AM

The toll system in the North East US is pretty good, although it doesn't deal with differing pricing based on time on the road, just distance traveled, and type of vehicle. Since it is almost exclusively used on individual long haul roads, or over individual bridges, rather than road networks it doesn't need to worry about time spent, distance traveled govern usage. It would, however, not be imposible to put up a similar system governed solely by time, tickets with entry time rather than exit number, much like a giant parking lot. A composite system covering both would be possible, although more difficult and of strictly limited utility: if time is a meaningful factor people are probably travelling around in circles enough that specific exit is nearly meaningless as a measurement of use, and if the road is long haul discouraging breaks by charging based on time produces a saftey hazard as drivers get tired.

EZ pass tags are mobile between vehicles, I believe photo's are taken when a vehicle passes through, in case the toll is challenged, to prove what vehicle passed through. The photo's are taken as much for driver protection by reducing fraud as for any tracking purpose, unless the driver speeds, runs the toll, or challenges the toll I doubt another person sees them. It functions much as a credit card, and I would hope the tags are issued ID #s such that multiple vendors could use tags on the same network. (which would reduce the effect of vendor wars.)
Although if the system was installed by a vendor it would be reasonable to give them a period of exclusive use as part payment for the instalation, and to assist with any debugging that needs to be done. I don't know if EZ pass is intended to ever allow multiple vendors.

Posted by: Michael on July 2, 2003 07:50 PM

The system in use in Australia does have multiple vendors. Each road operator using the system issues tags, charges people's credit cards, and then the tag is useable elsewhere in the system. There is a central clearing house that allows net amounts of money be transferred from one operator to another. (The operators are a mixture of governments and private road owners). Normally the operator that issues your tag pre-charges a certain amount to your credit card, and when that runs out they make another charge to your credit card. As you pay in advance, the different operators generally prefer it if you get the tag from them rather than their competitors, as that way they get free working capital from you.

Posted by: Michael Jennings on July 2, 2003 08:30 PM

A modest proposal, floating around my skull for many years:

Abolish Vehicle Excise Duty (it hasn’t been the Road Fund Licence since the 1940s).
Set a rate per mile, based on some semblance of reality of damage caused to the road. Monies to be paid into a Roads budget.
The vehicle’s owner pays as they go monthly for a percentage discount, or sticks it in a Building Society account.
Come the MOT, the odometer reading is used to determine the actual distance travelled. The owner has 14-28 days to settle up and gets a disk to display accordingly. Failure to pay and they start having to pay interest.

Advantages: anonymous to a degree, fair, road users pay for the bulk of road maintenance/construction.
Disadvantages: Having to get the odometer read in years 1-2 of car’s life, greater incentive to wind the clock back, and yes, the Treasury would have kittens.

Posted by: Adrian Ramsey on July 7, 2003 03:44 PM

I'm "this guy" mentioned above as concocting the anonymous electronic cash-in-buckets system. Obviously, the system is purely hypothetical, as well as being more difficult to set up than a license plate reading system, and much less likely to get you Homeland Security grants to set up than a total surveillance option.

However, there are (American) cities and counties out there interested in privacy concerns--my hometown of Ann Arbor, for example, is considering a resolution that would bar city law enforcement from participating in Patriot Act related investigations. It's not a resolution with many teeth, of course, as Patriot Act issues are mostly INS and FBI sorts of things, but it's the principle of the thing that they're going for. My point? Some cities would value the privacy of a cash-in-bucket scheme over the usefulness of the information gathered by a total surveillance system. And the only reason a TS solution can be "just switched on" and a CiB solution not is that the CiB solution isn't quite available yet.

Now, if only I could get the City of Ann Arbor to put their money where their mouth is and send some startup cash my way, I'd get to work on making the CiB solution as plug-n-play as the TS option is and provide you with a demo city.

Posted by: Murph on July 7, 2003 05:45 PM