Transparent Speed Cameras

According to the RTA website the number of motorists in the Sydney Harbour Tunnel that exceeded the speed limit by at least 10km/hr was 63% prior to the installation of speed cameras.  If we were democratic about such things then we should have concluded that the speed limit was at that time too low. More than 50% of motorists figured that a higher speed was more appropriate for their particular journey. The slightly increased chance of losing the rest of your life (due to driving fast) was obviously not as big a concern as the almost certain risk of losing a small part of your life (due to killing time in the slow lane). Obviously the concern of pedestrians wasn’t really a factor in this example.

Given the proliferation of speed cameras and electronic variable speed signage on motorways we ought to engage some technological and regulatory reform to at least make their operation and the application of speed limits somewhat more democratic. Firstly we ought to have devices that actually gather statistical data on the speed of cars along a given route rather than merely clicking pictures of those that drive above an arbitrary legal limit. And once we have that data we ought to use it to reform the setting of legal speed limits. Perhaps we could decree that if over a 3 month period more than 50% of motorists drive 5km/hr slower than the speed limit then the limit is too high and it should be reduced by 1 km/hr for the subsequent 3 month period. Whilst if less than 50% of people drive 5km/hr or more below the limit then we increase the speed limit by 1 km/hr for the subsequent 3 month period. In this way speed limits could be gradually reformed via direct democracy in action. We could call this reform “pushing the limit”.

And before you ask I should declare that I did indeed receive a speeding fine in the mail today.

26 thoughts on “Transparent Speed Cameras

  1. This is an issue that, in a way, goes to the heart of libertarian reasoning.

    Nobody could disagree that if speed limits are lower, fewer people will be killed and injured in traffic accidents. It is also obvious that to eliminate all road deaths, a speed limit of 5 km/h would do the trick.

    Obviously this would not be acceptable, even in subservient Australia. There is inevitably a trade-off between the convenience and economic value of faster transport and the inevitability of road deaths.

    The key question is, who decides what that trade-off ought to be – the government (which in practice means a faceless, nameless bureaucrat)or the people who are inconvenienced by slower speed limits or killed by higher speeds? This is not simply a matter of individuals risking their own safety either.

    Internationally there is a lot of literature on this subject. The solution is the 85th percentile rule (http://en.wikipedia.org/wiki/Speed_limit) Using this, the speed limit is set to the speed that separates the bottom 85% of vehicle speeds from the top 15%. The 85th percentile is slightly greater than a speed that is one standard deviation above the mean of a normal distribution. The normal distribution is based on speeds applying in the absence of speed limits.

    The theory is that traffic laws that reflect the behavior of the majority of motorists have better compliance than laws that arbitrarily criminalise the majority of motorists and encourage violations.

    I think it is a great example of libertarian thinking.

  2. during 18 years of driving in the UK, i amassed the grand total of one speeding fine.

    in my first three years in Oz, i have already been fined 7 times.

    it appears to me that the max speeds are deliberately set way too low and in effect act as a stealth local tax.

  3. 1. Personally, I could benefit from the advice of others when I have to decide how fast to drive. That is, I could benefit from the existence of speed signs. The people I’d like advice from are road engineers etc. who know more about the physics of moving vehicles than me. I’d also like advice from people whose job it is to analyse probabilities given that, like many others, my ability to ascertain and comprehend risks is far from perfect.

    2. As I’m sharing the road with others, and as many of them may share with me some of the above characteristics, I’d like those others to also have the benefit of speed signs.

    3. Should those speed signs have white or yellow background (ie legal limits or advice)? I’d suggest white, given that:
    (i) laws influence behaviour,
    (ii) I am likely to drive faster than the speed that is in the interests of both me and others, given that my imperfections outlined in para 1 would lead me to under-consider the costs of speeding, and given that I rank my desires (eg taking a risk to get there on time) over those of others,
    (iii) quite a few others are likely to be similar to me, and
    (iv) a person’s speed has a direct influence of the speed of neighbouring drivers.

    4. Given (ii), (iii) and (iv) above, and given the limits of the extent of (i) above it is probable that, on average, I and others will end up driving faster than what would be in most person’s interests, and above a speed limit set in a scientific (rather than with a ‘no-risk’, nanny state) fashion.

    5. This suggests a couple of policy prescriptions:
    (a) ensure a scientific rather than ‘no-risk’ approach to speed limits;
    (b) enforce the resulting speed limits thoroughly, through increased funding for traffic cops and speed cameras, given that the power of law to influence behaviour depends both on the penalty and the chance of being caught; and
    (c) do NOT increase speed limits simply in response to observations of average speeding.

    6. I suspect that increasing speed limits in response to observations of average speeding would serve to ratchet up speeds beyond the new limit (in the absence of any associated increase in enforcement or penalties). This would suggest that people set their speeds with some reference to the speed limits, as opposed to setting their speeds on the basis of some independent assessment of what an efficient speed is.

  4. A couple of points:

    1) The maximum safe speed differs from vehicle to vehicle and also depends on the driver. A porsche 911 can safely travel a lot faster than a 4 tonne truck can. That is because it is designed for it. It also depends a lot on the condition of the car – particular the wheel alignment and brakes. A car with poor wheel alignment can be dangerous at 100km/h whereas the same car with proper wheel alignment could safely travel at 130 km/h on a highway.

    2) Allowing citizens to decide speed limits is no different than allowing them to set the reserve bank exchange rate. A small percentage will know what they are talking about, the vast majority would just be guessing and you could end up with a much worse situation than you started with.

    3) Much of the money Duncan Spender wants to spend on speed enforcement would be much better spent on improving roads. The road toll is much more a result of poor roads than it is speed. That is why such a large proportion of deaths occur on country roads rather than city freeways. Median strips and crash barriers save a lot of lives, double marked white lines do not.

  5. Duncan – good of you to join us (for those who don’t know Duncan he helped found the LDP back around 2001). Duncan do you think engineers set speed limits? When they suggest limits should be higher do you think they don’t get ignored more often than when they suggest limits should be lower. I’m okay with science but on the basis of science Bob Brown wants to ban coal fired power stations within three years. Science should be in the hands of the people not a select few who rule over the people. This is really about direct democracy versus representative democracy not science versus ignorance.

    DavidL – I had LDP policy (ie implement the 85% rule) at the back of my mind when I wrote the article. However how do we get speed limits removed for long enough in order to measure what the speed limit ought to be according to the rule? The approach I’m toying with in this article seeks to side step that difficulty. Although perhaps my numbers need tweaking. For instance maybe in hindsight we can set the speed limit on motorways based on 30 minute interval data rather than 3 months of data.

  6. p.s. Duncan – in my Sydney Harbour tunnel example it seems that the mere existance of a legal speed limit didn’t stop most people from exceeding that limit. Enforcement is what slowed down the traffic. The RTA motto ought to be “getting you from A to B slower”.

  7. “it appears to me that the max speeds are deliberately set way too low”
    .
    If you look at accident and death rates in Australia per kilometer traveled, you’ll find some of the lowest rates in the world. Having not driven in Australia for about 10 years (before cameras and then after cameras), I was surprised at how nicely everyone drove once I started driving again. Before cameras appeared, everyone used to speed everywhere (including school zones and places like that — it was traditional to do 12ks over the limit). Thus whilst lots of people complain about the limits, and no doubt some are wrong, there is a flip side to it. It seems to me one of the main groups forced into behavior change because of it were young dangerous males — now all they can do is accelerate up to 60ks, and that’s hardly a bad thing.

  8. Conrad – we could make it safer still by halving the speed limit. It isn’t about whether slower is safer, most people accept that, it is about where you draw the line and who decides. I like direct democracy via revealed preferences (somewhat like the free market) whilst others prefer supposedly wise leaders that command and control.

  9. The roads also pay a part in this some highway sections may only be able to handle a speed of 140 Kmh, a friend working in DoT Queensland mentioned this. The poorer sections of highway much less, so even if your car is a Porsche the road may not be safe anyway. I listened to a radio program about speed limits and the statistics it presented said that every 5 Kmh over the speed limit doubled your chance of an accident.

  10. “I listened to a radio program about speed limits and the statistics it presented said that every 5 Kmh over the speed limit doubled your chance of an accident.”

    I may be interpreting this wrong or it is being deceptive (somebody can correct me if I am), but I don’t know how they come to these statistics. They can’t measure how many people are speeding and not having an accident because they never become a statistic.

  11. As a minarchic libertarian, there is nothing wrong with the owner of the property, the RTA, setting conditions on its’ use.
    However, a city in Germany recently minimised their displays of speed limits, and let drivers set their own speeds in many areas. Without an ‘official’ limit, and thus no guide to what was a ‘safe’ speed, drivers were more cautious, and they had less accidents, and the city also saved money by not needing to replace speed signs.
    Mark Steyn, writing in ‘The Spectator’ magazine, also mentioned a Dutch bridge without a speed limit, which also had very few accidents.
    so it seems that having a limit to judge yourself against, is the thing that encourages speeding.

  12. “it is about where you draw the line and who decides”
    .
    I imagine you could do that for lots of things if you wanted. Land boundaries, class sizes in schools, immigration, etc. all come to mind.

  13. For 50 and/or 60 zones there’s far too many to ever police speed on that effectively, so why not just limit police powers there to “reckless driving” laws. In school zones, traffic lights (with red light cameras) for kids to cross at should prove sufficient, and is probably more cost effective in the long term than cross walk people.
    For more major roads, private companies (lots of them) can make a profit at a competitive price (with a discount for locals, perhaps) and with a more long term vested interest in their reputation than politicians, upkeep roads and take measures to increase safety much better than governments, which don’t even seem to have a coherent idea how they want to pay for roads.

  14. The only way going 5kph over the speed limit would double your chance of having an accident is if the speed limit in question was about 13kph. That’s the only example where the kinetic energy involved would basically double with an extra 5kph, in which case breaking would be twice as hard.
    By the way, Terje, if you want to make a facebook group advocating this, and havn’t already, I’d totally join. Try avoiding words like reform though. Most people are turned off by that. But something like “set speed limit at what people actually drive” would have quite a bit of mass appeal, methinks. Or I could do it…

  15. Surely doubling kinetic energy and doubling your chance of an accident are two different things. Reaction time is probably more important here.

  16. If anything, alertness is increased in more dangerous situations, such as faster driving, in which case the mental reaction would be quicker. (Not that anything you said contradicts that; just saying)

  17. A speed limit set too low could also encourage infringements and risky driving behaviour as young men try to overtake the car with the hat on the parcel shelf. The safest speed must surely be the speed that supports safe driving and prompt arrivals.

    I’d like to see an analysis of whether then change to 50ks in the suburbs has a measurable impact on accident statistics.

  18. Speed limits are definitely too low, given how much car safety has improved in the last decade.

    I agree with fine-tuning major roads and freeways.

    Adjusting them by multiples of 1kmh or 5kmh seems too fine. Make adjustments multiples of 10km and round upwards.
    e.g If the average speed is 72 or 75, then make the limit 80.

    Leave the suburbs as a standard 60kmh zone to avoid confusion, and perhaps 50kmh speed limits for side streets and school zones.

  19. Speaking of road safety, has anyone noticed that the assumed average (rather than ideal) reaction and response time in Australian state government road safety literature (and advertisments) is 1.5 seconds. Which is mirrored by various real world studies on driver reaction times.

    But a reaction time of only 1 seconds is assumed for traffic lights! (see Austroads Guide to Traffic Engineering Practice)

    This is not a problem on flat intersections however, because the standard amber interval (4 seconds in a 60 km/h zone) caters for a grade of up to 5%.
    However what this really means is that traffic signals on roads with significant (around 2%+) grade are not properly engineered!

    This can be the difference between a heavy vehicle stopping in time, or going through a red light or entering the intersection when the perpendicular lanes have a green signal!

    Secondly curves directly before intersections are not factored in either, unless there is a lack of visibility. Despite the fact that it can be dangerous to brake hard on bends in wet or icy conditions, especially for vehicles towing trailers.

  20. The danger is in increased reaction and braking time (at low speeds) and the possibility of losing control of – or rolling – the car while turning (at high speeds). Both of these are of course subject to what kind of car you are driving as I said earlier (much easier to roll a high centre of gravity Kombi van than a low-slung sports car).

    Its not simply a matter of kinetic energy or momentum. That will only affect the magnitude of force applied in the accident, not the possibility of an accident occuring.

  21. @Mitch Comment 17.

    Sure your reaction time improves when under certain stress but what happens when the excessive speed is routine? when you travel the same stretch of road, day and day out 15 Kmh faster than than normal and the unexpected happens.

    There will be no improved reaction time, just a nasty accident.

  22. Here’s my theory, I’d be interested to see what people thought:

    It seems fairly self evident to me that it is safer when everyone is going roughly the same speed.

    Because if two cars travelling more or less the same direction at speed X hit each other via side swiping, the forces are similar and the impact won’t be as severe as if there was a large deceleration/acceleration force also involved (that would result due to a speed difference).

    I think driving in 3rd world countries where there are a large variety of vehicles on the road is dangerous partly because of the difference in vehicle speeds. The difference in speed makes judging other vehicles more difficult especially oncoming vehicles and can result in more unpredictability from other drivers due to the difference in time taken to perform certain manoeuvres like turning, braking, accelerating, U turns.

    This idea, I think would support the 85% rule because the 85% rule is trying to determine the speed most people feel comfortable driving at, therefore this should make it easier for people to drive at similar speeds.

    Incidentally I also think that the way the government advertises for going slower shows appallingly bad logic and reasoning.

  23. People might be more obedient if signs were more polite. a sign that said “PLEASE don’t kill yourself- drive at less than eighty if you want to be safe!” would get a more favourable reception than an arrogant- “don’t drive above 80!”
    I wonder if this should be more universally applied? “Please don’t murder anyone- you might get caught and put in jail! (And miss out on that big date on the weekend!)”
    “Please don’t take drugs! You might not like them, and we might have to arrest you and put you in our over-crowded prisons, full of hetero- and homo- sexual rapists!”

  24. It occurs to me that our laws should be designed such that they protect our lives and possessions. With speed limits (and for that matter drink-driving, motorcycle helmets, seatbelts etc) we are being punished for an offence against somebodies life or property without having caused any real damage. How about we abolish them and replace them with swift, severe, very public punishment for when an actual offence is committed?
    The basis for these laws is severely flawed. Using the same logic you may as well charge all accountants with embezzlement, retail employees with theft and spouses with murder on the basis that theses groups are statistically more likely to commit these offences.
    Lets hear it for personal freedom and (this is the important bit) personal responsibility.

Comments are closed.