% The Flaming Right by paul murphy

Alberta safe speed zones

Nearly everybody drives, and nearly everybody has an opinion on both traffic law and traffic law enforcement, but as a society we don't know what works, or what doesn't, because the answers aren't obvious; because police, journalists, and government all line up to favor more enforcement; and because access to the data needed to understand what's happening is largely restricted to those representing regulatory and enforcement agencies with regulatory and enforcement agendas.

The use of speed cameras is, for example, controversial with a majority of drivers seeing them mainly as road side tax collectors, while politicians and other system insiders endlessly assure us that the revenue produced is immaterial relative to the safety benefits conferred by their use.

The money side of all this ticket writing seems obvious, but isn't. The revenue numbers are not generally available, but a reasonable estimate is that Alberta's 2.9 million or so licensed drivers will get about 5 million tickets, and pay about $950 million in fines, during fiscal 2016/17 - costs aren't public either, but include obligations like those in infrastructure support, pensions, and data processing undertaken to support increases in enforcement but unlikely to be reduced with it, and you'll guess that costs come to somewhere around 60 cents per revenue dollar for fiscal 2016/17.

The financial effects on municipal government aren't clear either. The use of red light and speed on green cameras, for example, is very effective for raising money but is also widely known to be closely correlated with increases in minor accidents and those increases are forcing municipal governments to spend far more than previously thought necessary on left turn lanes and signals. Once those lanes are installed, however, the number of tickets written drops and cities generally make their problems worse while moving ever further from the safety rational by moving the camera in 90 degree increments around the intersection to protect the monies sunk into the senor grid.

Take the bigger picture approach, however, and it's obvious that ticketing is a net loser for government as a whole if it is effective in reducing average vehicle speeds - because the total tax content (federal and provincial) per dollar in retail sales for fuel produced in Canada now runs over 70 cents and even a 5% increase in average vehicle speed would produce incremental revenue without new costs across all levels of government significantly exceeding what the province and its municipalities net from ticketing.

Thus about the only thing that's actually clear about the revenue argument for ticketing is that the biggest beneficiaries of Alberta's ticketing extravaganza aren't our municipal governments and police services, but the companies that write the camera tickets and out of province insurers - because most of the tickets issued by police or sheriffs today are more about bad luck and nearly continuous surveilance than about habitual dangerous driving but still incur the demerits that formerly justified higher premiums.

An aside:

One of the consequences of today's ticketing extravaganza is that parents facing the effective certainty of significantly increased ticket and insurance costs for their teenaged drivers are telling the kids that those costs are on them - with the result that many are simply refusing to get licensing while still at home. In the short term this downloads much of their transportation costs on the parents and those of their freinds who do get licenses but, in the longer term it means that these kids are being denied the opportunity to practice driving while under parental supervision and will ultimately be involved in far more fatality and injury incidents than will those who do get early training and support.

More importantly, however, the real costs are not financial but human and social because the system treats everyone badly while weakening the respect for law our society is based on. Normal six year olds don't, for example, want to become bureaucrats, but the manpower intensive nature of government means that thousands of people who might otherwise achieve something with their lives waste years working in the collection system instead; police officers have their jobs diminished; clogged courts and over worked prosecutors allocate scarce resources to non criminal matters; the frequently caught become immunized to the operation of law; good drivers trade greater driving costs and risks for reductions in ticketing risk; and, all road users are exposed to ever increasing road and legal risk as more and more people, forced to drive outside the framework of law and responsibility (unlicensed, unregistered, and uninsured) rationalize increasingly dangerous behavior to avoid being caught.

The safety side is even more opaque in part because the marginal effect of a change in enforcement on road safety is obscured by dozens of other factors; in part because seemingly positive short term change can mask negative longer term effects; and in part because those who defend the system present a partial and one sided view to the public while blocking outsider access to real data.

Three examples:

  1. it is a classic of the pro-camera argument to cite an intersection or piece of roadway where accidents went down after speed, red light, or speed-on-green cameras were installed while hiding the twin realities that "solved" problems were often just statistical anomalies, and that regional totals usually increase with camera use.

  2. Alberta's camera and other enforcement enthusiasts point out, correctly, that traffic death rates have, at least until the sheriffs really got going in 2011/12, been going down in Alberta. However, in claiming at least partial credit for this, they don't usually mention:

    The most likely thing, of course, is that there is a point beyond which increases in enforcement activity make roads more dangerous rather than safer - but the numbers needed to disentangle the effects of Alberta's increased enforcement efforts from those of other factors such as increasing urbanization, population aging, better roads, and better vehicles are kept from the unwashed.

  3. politicians, journalists, and police spokesmen defending enforcement generally deny that the continual threat posed drivers by enforcement efforts like sheriffs and cameras negatively affects driver performance by saying that people who don't speed have nothing to fear - but the same people cite driver performance to support demands for stringent penalties against drivers who voluntarily expose themselves to other distractions. Most directly, however, adaptive behavior by the frequently ticketed exposes everybody else to greater road risk - for example, the platooning behavior we see as people pile up behind drivers obsequiously obeying every speed limit is stupidly dangerous.

    More subtly, the claim that people who obey the law have nothing to fear directly shows both the intent to intimidate and the reality that people respond to fear - thus a driver pulling out to pass on a two lane highway will be distracted by the reality that any hill, copse, or ditch could be hiding a sheriff. This fear is intended to affect behaviour, and does - but not always in positive ways. Thus it may well be true more often than not that the bear won't charge, the mugger won't shoot, and the cop won't ticket if the person facing the threat does nothing to trigger the reaction, but placing drivers under continual surveilance stress cannot make our roads safer - on the contrary it distracts driver attention while breeding alienation, loathing, and contempt: feelings the courts, which assume the ticket writer truthful and the accused therefore guilty, reinforce because they offer no effective recourse to those without the money or connections needed to get charges withdrawn, stalled, or dismissed.

All enforcement is, of course, intended to affect driver behavior - unfortunately, inappropriate enforcement leads to counter-productive adaptation. Consider four examples, all from Lethbridge but all with correlates across Alberta:

Whoop Up Drive seen from the West

  1. Whoop-Up drive carries about 26,000 vehicles per day each way across the Oldman river. The road is posted at 90Kph at the crests, and the city routinely hides mobile speed cameras just above the point where the road turns and flattens before reaching the bridges - enabling them to photograph traffic about where vehicles using only engine braking typically reach their maximum speeds.

    As a result people who regularly drive this route have become so habituated to heavy braking just before reaching those positions that all lanes have dips in the pavement where this happens - and if conditions happen to be bad, or somebody's tires a little slick, that can have disasterous consequences. A surprise October snow dump on a few years ago led, for example, to a number of separate, multi-vehicle, incidents in both directions involving, in total, more than 40 vehicles and at least a half dozen injuries.

    Two asides:

    1. The city has recently responded to the crashes on whoop-up by reducing the speed limit when they believe conditions warrant it. This has produced a reduction in the severity of each incident at the cost of creating more incidents typically involving far more vehicles. The reason this happened is that reducing average vehicle speed increases both traffic density (decreasing distances between vehicles) and the effects of disparities between the braking performance of better and worse performing (with respect to braking under the conditions extant) driver/vehicle combinations.

    2. These cameras are something of an environmental disaster: not only does the heavy braking just where vehicle speed is naturally maximimized erode the pavement while spewing brake and tire particulates, but most vehicles using their brakes to stay at 90 near the end of the downhill then burn an additional 1/15th of a liter or more to maintain that speed going up the far side - and this doesn't sound like much but makes a big number (about 1.25 million liters) when annualized for 52,000 vehicles per day.

  2. you see strange driver behavior at interesections widely thought to have hidden cameras.

    The problem is pernicious: intersections featuring speed on green cameras have a trap zone whose position varies with conditions - particularly road and vehicle conditions.

    For every combination of vehicle, road conditions, approach speed, and amber duration every controlled intersection has a piece of roadway on which, if the light turns amber while the driver is in it, all of the driver's choices are bad. For example, someone approaching a controlled intersection with a 3 second amber phase at 13.8 M/sec (50Km/hr) in a vehicle capable of a maximum braking rate under conditions then extant of 4 M/Sec/Sec will enter the intersection on amber if he's 41 meters from the sensor line when the light changes and does nothing in response but will get an expensive ticket in the mail a month or so later if he's 42 meters away when this happens and he does not respond.

    Thus a driver who is 42 meters away when the light changes has no safe options: continuing at 50km/hr gets him a ticket for entering on red; speeding up to clear the light earns a speed on green ticket; if he takes 1.5 seconds to decide he's not going to make it and eases on the brakes, he'll slide into the intersection and get an expensive reminder in the mail long after he's forgotten the incident; but, if he either has fast reactions or really slams on the brakes, he will avoid the ticket, but is likely to get a rather forcible introduction to the driver behind him - especially if the latter is a bit slow to react or in charge of a vehicle, like a heavy truck, that can't match even average braking performance.

    Watch traffic at any intersection where people believe traffic cameras to be hidden and you'll see the result: dangerous adaptive behavior as people substitute driving risk for ticket risk. People will, for example, slow to 30, 50 meters or more ahead of the light and then accellerate hard when it's still green as they get closer; switch rapidly (and often without warning) into the left or right turning lanes when the light changes; floor it just before the crosswalk; or, trigger chained panic stops in which the trapped driver first eases on the brakes but then panic stops short of the crosswalk while the next in line does much the same thing, and so on.

    An aside:

    Some people object to the use of "trap" in this context. However, it seems fair to assume an intent to trap the unwary when most cameras, and camera vehicles, seen on the streets are disguised or hidden and police make statements like this one attributed to Police Sgt. Joerg Gottschling the Calgary Herald by the Calgary Herald:

    "Nothing else has changed," he said. "And if you see a Ford F150 pulled over on the side of the road with its hazard lights on . . . it might be a good idea to slow down."

    because the statement is absolutely true, but utterly misleading in that very few, if any, of the camera vehicles seen on the streets have their hazard lights on.

  3. Similarly, the fact that the mobile camera operations are rarely seen near the centers of school and playground zones produces dangerous adaptations.

    Most importantly what you see at almost every school zone here is people speeding up immediately after passing out of the range of the camera vehicle parked near their point of entry to the school zone - meaning that their primary focus is on the threat the cameras pose to them, not on the threat they pose to school children.

    Less commonly it works the other way: people blasting through the school zone, slowing to 30 near the end to pass the camera, and then accellerating hard as soon as they pass out of range.

    Laval school zone from the south The south facing van in the photo is about twenty yards from where the southbound end of the school zone was at the time the photo was taken.

    Since then the city has moved its school zone signs about a 100 meters closer to the last crosswalk near the school. This has made things safer in part because people are less likely to disregard a school zone that's actually close to a school, but mainly because people slamming on the gas upon leaving the school zone now have about a 100 meters in which to regain control before entering the downhill curve in the foreground - where dozens of people doing this have, over the years, smashed into houses, trees, and each other.

  4. When the province finalized the August 2011 renewal of its policing agreement with the federal government both its use of Alberta court sheriffs as highway patrol officers and the Police College development in Fort McLeod became redundant, but only the college was cancelled - presumably because the highway patrol was seen as producing $70 million in revenue from half a million tickets during fiscal 2012/13.

    Roughly one third of all accident investigations in Alberta report "following too closely" as a contributing cause - and the key effect the sheriff operation has had on traffic has been to greatly increase plattooning on secondary highways as people with too many demerits drive at just below the limit with one eye on the speedometer, one eye scanning for sheriffs, and one eye on the tailgaters grouping up behind.

    People from Lethbridge are now, for example, taking a longer route to and from Calgary because enforcement activity on the shortcut using highways 23 and 519 has made doing the stupid thing smarter. Basically what's happened is that zealous enforcement has turned what used to be a relaxing drive in which drivers simply passed the heavy trucks and occasional slowpoke favoring this route into a white knuckle experience in which passing slower vehicles slowly risks head-on collisions, passing them safely risks getting ticketed, and hanging back increases the risks to everyone as long columns of stressed and frustrated drivers build up behind.

    Since this started, in 2012, dangerous passing behavior on those highways has been killing one or two people every year - compared to none in the years before.

    (As a black humor aside, a traffic investigator told me that two motorcyclists on Highway 23 killed each other during an attempt to pass a line of vehicles when their LIDAR threat receivers went off and the leader braked sooner and harder than the other while veering right. Not exactly a common modality, but devastating if true.)

The most obvious result of all the enforcement activity is a shifting of responsibility for driving decisions from the driver (who sees the state as enforcing its judgment over his regardless of conditions) to the state and thus an ever increasing commitment to driving to rule - meaning, for example, that people will push speed limits regardless of conditions and are becoming both increasingly less likely to yeild right of way and more likely to maintain speed crossing intersections, like those in which one rural highway cross another but only one has stop signs, where it is legal but stupid to do so.

Similarly, some rules are simply counter-productive so enforcing them directly produces injuries and deaths. In 2014, for example, police started using aircraft to detect and ticket people driving in the "fast" lane on Alberta's highways. As a result more people are now driving, particularly on empty highways and at night, in the "slow" lane - where they're hitting significantly more pedestrians, deer, cyclists, and stopped or slow moving vehicles than ever before.

In Alberta, as elsewhere, growth in enforcement has been driven by four main factors:

  1. The first one is a particularly vicious spiral in which increased enforcement leads to increases in the number of reported incidents, and those increases are then used to justify further increases in enforcement.

    School zone cameras provide a classic example of this with every new school zone incident producing more demand for more enforcement - and more enforcement producing more incidents. In support of this cycle police and approved researchers usually claim that the negative correlation between tickets received and accidents (in general it appears that the people who get the most school zone tickets have the fewest accidents) proves enforcement works; but the opposite is actually true. The reason for this is that dangerous drivers avoid cameras while safe drivers avoid kids - so the tickets go to the people least likely to be involved in accidents and both the number of incidents and their relative severity rise with enforcement.

  2. The second is the political analog to the first: people enjoy speed but under-estimate collision and injury risk, so if enforcement makes speeding on wide roads designed for higher speeds seem too risky, they'll do it in residential neighborhoods and on backroads where the physical risks are a lot higher, but the chances of getting a ticket are a lot lower. That, in turn, produces significant political support for more enforcement as residents respond to the obviously irresponsible behavior they see in front of their homes by demanding something be done to stop it - a demand which is interpreted as favoring more enforcement, and the enforcement focus on busier roads than makes the problem worse.

  3. The third reflects the operation of a classic moral hazard: the public is told that enforcement is about safety, but the sheriffs were internally justified on turning a divisional profit, a drop in oil revenues triggered a doubling of traffic fines (effectively lowering speed limtis by making the prosecution of lesser offences profitable), and the presentations the people selling camera services to municipalities make behind closed doors generally include thinly disguised briefings on what to say to reporters, but are otherwise entirely financial in focus.

    Add the fact that nobody really knows what works for, or what works against, traffic safety and the combination of financial and political presure explains why enforcement managers facing this particular moral hazard almost invariably make the financial choice - favoring revenue over safety by allocating enforcement resources on traffic density, not traffic risk.

    The net effect, however, is that traffic cameras and other speed traps are usually put in places where the majority of drivers think the posted limits are too low simply because that is where they affect the greatest number of drivers and are therefore most profitable, while the majority of the dangerous behaviors these measures are nominally intended to reduce take place in areas where police, and cameras, are rarely seen.

  4. frustration with the system's inability to cope with minor crime in general and serious crime committed by judicially protected minorities in particular, leads police officers and others involved on the enforcement side to focus on areas where their efforts are rewarded. As a result, while it's true that most officers do not like writing speeding tickets, they'll spend time on it because it feels like they're doing something - and the system wide performance metrics on which they're judged provides positive feedback for doing it. Thus someone caught doing 80km/hr in a 70 zone on a nice dry four lane highway in the middle of nowhere (e.g. on Highway 16 just inside the Jasper Park boundary or westbound on Highway 3 outside Coaldale) can draw multiple cruiser attention - but native vandalism resulting in a dozen damaged vehicles in West Lethbridge isn't worth a minute of police time.

The bottom line on all of this is that the government's requirement that drivers cede judgment to them while also holding them responsible for the consequences is untenable without some form of social agreement on where the limits to both are - meaning most directly that today's exponential growth in enforcement is unsustainable in the long run and, indirectly, that jurisdictions continuing down this particular path to totalitarianism will eventually find their economies crippled and their people moving out.

This doesn't have to happen - four ideas toward some solutions:

  1. Canadians are endlessly assured, and widely believe, that our court system treats everyone equally and assumes defendents innocent until proven guilty. In reality, however, defendents in traffic cases are assumed guilty by the courts, the well connected have charges dropped before things get to court, people with the right skin colors and dependencies get penalties waived, lawyers for the well off or nearly rich can often manipulate the system to get charges stalled or dismissed, and thousands of ordinary people go to court expecting a fair hearing but only one or two a year get one, and only the very rich and very angry can force courts to fully consider the facts.

    Those who favor enforcement claim the process isn't about money, but their defense for the existing system is entirely financial - it simply isn't cost effective, they say, to give everybody a fair hearing. As a result Alberta has been evolving an interesting system: one which takes contempt of process to new heights as harried prosecutors and hearing officers simply reduce charges for those willing to stand in line to see them and then plead guilty without a trial.

    As it stands the system is grossly unfair to almost everybody - and that has to change. Some suggestions:

    1. require video and technical evidence, including relevant equipment and sensor logs, for all traffic charges; and deliver that evidence to the accused within 24 hours of the alleged offence;

    2. make all charges, and their disposition, public;

    3. end plea bargaining;

    4. end concurrent and time-served sentencing for traffic related offences;

    5. require traffic cases to be heard by special traffic juries of six, give those juries nullification authority, have each juror sit for several weeks and thus hear many cases, draw them from among qualified drivers who are retired and over some age like seventy; and, where possible and applicable, adjust schedules so that officers who are called on to testify in many cases over one jury period do so in front of the same jury members;

      These juries, and the political, legal, and media policies supporting them, have to take the nullification option very seriously because this is the remedy for absurdities like ticketing people for speeding while passing, not coming to a full stop when turning on a red with no cross traffic and an idiot coming up behind, etc. Basically, these juries replace the good sense and judgment police were expected, in the 1950s, to apply in ticketing - and their decisions in throwing many of these tickets out will eventually affect enforcement actions in positive ways.

    6. require judges to justify, in writing, any changes they make from specified penalties in cases where the jury has reached a guilty verdict - and make those judgements available, without redaction, to the public; and,

    7. require the crown to notify and compensate people who paid tickets for offences that were subsequently proven not to have been offences.

      A personal example: I got a ticket for doing 71 in a 60 zone at the Scenic drive exit from Whoop-up and paid it because this happened going up a steep hill where the bottom is posted at 90 and a point just below the top at 60, so I assumed the ticket was legitimate. As it turned out, however, they ticketed many people for this on that day, apparently including one with insider access who promptly had the ticket withdrawn because the 60 zone didn't start until well past the section visible from where the LIDAR operator had been hiding.

      The next week a city crew moved the speed limit signage one lamp post down the hill to put that piece of road within the 60 zone - but I do not believe a single person who either had already paid the ticket or did so subsequently was ever contacted, compensated, or had the wrongfully assessed demerits occassioned by these tickets removed.

      More to the point for this essay, the city then put an electronic speed monitor near the top warning anyone going over 60 - in bright flashing red - that they might be ticketed. Unfortunately, traffic on Scenic isn't visible until you're past that point, the merge area at the top is short and narrow, and the platoons released by the traffic light just north of that merge on Scenic move at 60Km/hr too - so now instead of sliding into the traffic stream on Scenic and then slowing to match traffic, the unlucky are trapped between braking hard on entering the merge area or trying to squeeze in where they shouldn't, and what used to be a fast and safe junction now features frequent collisions.

  2. In many cases the combination posed by today's drivers, vehicles, roads, and traffic conditions differ quite radically from the assumptions underlying roadway engineering recommendations on signage, passing, speed, signal durations, and many other factors affecting the safe operation of our roadway traffic system - and, of course, the actuality often combines those recommendations with years of politically driven change, usually downward with respect to speed limits and almost never in response to engineering issues.

    As a result driver opinion on what is, or is not, a safe speed under the real world circumstances defined by the road, the car, the driver, and weather conditions can vary dramatically from posted limits.

    The recommendation is therefore to find ways to take driver judgement into account in setting speed limits.

    One idea would be to experiment with "right foot voting" for speed limits on some roads - on, for example, major provincial highways or city bypass routes. Under right foot voting, the speed limit would be continually revised according to how most people actually drive. One approach, for example, would be to set the limit on a road segment to the higher of an engineering default or the speed under which 97.5% of the vehicles on that road segment over the last thirty days during which weather and road conditions were comparable were driven.

  3. Police spokesmen and other enforcement enthusiasts always claim that enforcement is about safety - and it should be true, but it isn't.

    For example, the total of all Alberta collision reports for highway 22 south of Longview in the last thirty years before the sherrifs were deployed in which injuries were suffered, speeding is cited as the major reason the accident happened, and no unusual driver, vehicle, or road conditions were reported is zero or very close to it. Exceed posted limits by even a little bit on that highway now, however, and your chances of getting ticketed as a clear menace to society are very good - and the number of injury crashes, including some fatalities, has increased in close correlation with the increases in enforcement.

    During that same thirty years roughly one third of all Alberta collision reports list following too closely as a significant contributory cause - but your chances of getting ticketed for viciously tailgating somebody toddling along at 49 in a 50 zone with one eye on you, one eye on the speedometer, and one scanning for cameras, vary from slim to none.

    The reason the system prefers issuing speeding tickets to going after irresponsible behavior like tailgating is that the former produces both easy money and easy numbers for the annual performance report, while the latter entails risk, complexity, court time and small numbers for both cops and prosecutors - all with little offsetting revenue for the system.

    The recommendation, therefore, is to re-examine enforcement methods and priorities to put safety ahead of revenue - to put, in other words, enforcement resources toward efforts that may do good while ignoring low risk behaviors.

  4. As presently constituted the entire system casts police, crowns, and the judiciary in disrespect; forces an us-against-them mentality on all involved; and, gives everyone on the enforcement side (Police, Crowns, Judges) clear financial and professional incentives to band together in securing and enforcing convictions.

    To at least partially address this problem traffic enforcement should be rigidly separated from general law enforcement. It does not take a lot of training to write tickets, highly trained police officers should not be wasting their time doing it - and the incentives should drive an adversarial relationship between those who write tickets and those who enforce them.

    This can be achieved if two programs are implemented.

    1. first, all ticketing (including camera operations and checkstops) should be turned over to a provincial "peace officer" force of dedicated ticket writers with "trusted observer" access to police and an obligation to act in a police support role when asked to do so by police, but who are not themselves police and do not in any way pretend to be police.

      Since ticket falsification is quite literally highway robbery, the ticketee's first recourse would then be to a panel consisting of the ticket writer's immediate supervisor acting in an advisory role to a trained police officer who will determine whether, and against whom, charges are laid,

    2. and, second, the financial incentives to ticketing should be removed along with the financial penalities to protesting tickets. To do that all ticketing proceeds (gross, not net) should be distributed among the acquitted (pro-rated per dollar of the fines initially required of them) within four months after the end of each fiscal year.

The bottom line on Alberta traffic enforcement is that not knowing what works, and what doesn't, in terms of meeting the road safety agenda has led to a system that's out of control, secretive, and almost entirely about the money. Fixing it means opening the data to the public, finding ways to make the court process fair, and making enforcement about safety, not money.

The general policy prescription is therefore the obvious one: we need to find out what works and what doesn't so we can do more of the former and less of the latter - and that requires two things:

  1. enforcement needs to be about the law, not money - and the law needs to have a basis in actual safety research, not uninformed belief and political horse trading; and,

  2. all data collected by government, including traffic and driver data, must be public by default and confidential only if genuinely necessary for purposes other than the protection of government employees and decision makers.


Paul Murphy, a Canadian, wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.