6

BMW drivers are widely thought to be poor drivers. For example, in a poll for Auto Trader:

Around 21,000 motorists were quizzed, and 45 per cent said BMW drivers were the worst on the road.

Is this perception accurate? Are BMW owners more likely than owners of other cars to be involved in accidents or more likely to have convictions for driving offenses?

Tom77
  • 11,605
  • 8
  • 61
  • 87
  • 5
    @Tom77 - It is simply an attack upon people who are perceived as privileged. For some reason this is acceptable even though if it were replaced with a minority group or the underclass it would be seen as the ignorant bigotry that it truly is. – Chad Jan 18 '12 at 15:13
  • 2
    Naw! The Top Gear guys tell us that this title has passed to Audi drivers... – dmckee --- ex-moderator kitten Jan 21 '12 at 18:09
  • 4
    "worst on the road" according to poll, pretty much means **the most annoying**. I doubt it has anything to do with measurable safety metrics. – vartec Jan 23 '12 at 10:14

4 Answers4

13

BMW drivers are not the worst on the road:

I found a comparison of accident rates for drivers of different cars.

The website looked at over almost a million car insurance quotes in order to produce the analysis. The data looks at different car manufacturers and car colours and correlates these with claimed driver accident rates in the five year period before the quote was run.

table

BMW drivers appear to be more likely to have had an accident (16.7% of BMW drivers reported that they had an accident in the previous 5 years compared to 13.3% for all drivers), but they are not the most accident-prone drivers.

Tom77
  • 11,605
  • 8
  • 61
  • 87
  • Interesting, that would support risk compensation theories, top ones (with least accidents) are notorious for not providing lot of safety features, while few of the bottom ones are famous for this (like Saab or Volvo). – vartec May 03 '12 at 14:08
  • I can not believe red cars are the least likely to be involved in accidents. – kotekzot May 04 '12 at 01:19
  • 2
    @kotekzot - Red *and* yellow cars are least likely, according to this. I note that these are generally the most visible colors as well, although that depends on the shade. I believe my next car will be safety orange with blinking, dancing polka dots. – Goodbye Stack Exchange May 04 '12 at 02:28
  • @neilfein yeah, I also thought about visibility, but how unfair is it that sensible grey car drivers are 43% more likely to be involved in accidents than macho red car drivers? All the more proof that we live in an indifferent universe. – kotekzot May 04 '12 at 02:40
  • 3
    It appears that the claim **is true if the BMW is white** (excluding those maniacs in grey Lexuses (Lexii?)). So if you own a white BMW, best to spray it red asap. – Reinstate Monica - Goodbye SE May 04 '12 at 11:35
  • 4
    White BMWs are driven by the Police in the UK, maybe it is true. – davidjwest May 04 '12 at 21:16
  • What if it were the vehicle rather than the driver? – Paul May 05 '12 at 03:22
  • 2
    This data is probably skewed by the fact that people with fancy expensive cars are more likely to report any little accident they have to their insurance company, tallying one more into their statistic, whereas people with one of the less awesome cars above might not make a claim on the same accident and just hammer it out on their own. People with cheap cars are also more likely to opt for high premiums because they can't afford high rates, which means they're much more likely to foot the bills themselves when an accident happens. Insurance companies won't find out about these accidents. – Alain May 06 '12 at 14:30
  • Anyone else notice a complete lack of sample sizes and error bars? – Oddthinking Sep 26 '13 at 07:22
  • @Alain: "People with cheap cars are also more likely to opt for high premiums" - I think you mean "high excess amounts" – Oddthinking Sep 26 '13 at 11:27
  • 2
    @Oddthinking I suspect there's a large variance in sample sizes, e.g. in 2012, there seem to have been [33x as many Fords sold as Lexuses](http://www.carmagazine.co.uk/News/Search-Results/Industry-News/UK-2012-car-sales-analysis-winners-and-losers/), so I wouldn't be suprised if some of the make / colour combinations have really low sample sizes. The relatively large spread also suggests this, but that's just an opinion. Combined with a variety of other factors, such as likelyhood of reporting, kms driven on average by brand, etc, I'm not sure I'd put much weight into this source. – Daniel B Sep 26 '13 at 12:23
  • @Daniel B: Agreed. Look at how little difference colour makes for Fords and Toyota compared to Saabs and BMWs. Maybe that's natural variance in low sample sizes. Meanwhile, I am saving up for a red Porsche - they never have accidents. – Oddthinking Sep 26 '13 at 13:02
  • @Oddthinking that's my feeling too, the numbers are just so diverse that it's more likely to be noise due to small samples. That said, this probably reflects the kind of "logic" insurance companies apply when determining our premiums... It would actually be interesting to play with a good data set on this topic, e.g. including number and severity of accidents per km driven, avg. kms per year per brand, value of repairs vs. cost of vehicle, and so on, but I'm afraid this isn't it. – Daniel B Sep 26 '13 at 13:22
  • @DanielB small sample size is one factor, but is driven by popularity of a brand/type by the target audience of a specific brand or type of car. E.g. here we see a lot of black VW Golf TDIs and Peugeot 307 diesels being driven by 25-30 year old guys as their first lease car, traditionally a group that has a high accident rate (it's their first car typically, it's a company car, they don't suffer financially for damage to it, and they're still young hotheads too, a deadly combination). – jwenting Sep 26 '13 at 13:29
  • 2
    @jwenting I do get that, and I'm sure there's interesting things to be learned along those lines. That said, it's strange that patterns in colour are so marked (up to 10x more accidents), yet vary among brands (e.g. contrast Daweoo and Fiat yellow vs red, likely involving a typo). The biggest of the "surprising results" seem to coincide with brands and colours which seem to be rare(er), which could easily mean issues with small sample size. In the absence of evidence to the contrary (confidence intervals, samples sizes, etc) I don't think this chart carries that much weight. – Daniel B Sep 26 '13 at 13:40
  • Yay - I have a black Subaru...I'm going to go argue my case for a lower Insurance Premium citing this site :-) – Rory Alsop Sep 26 '13 at 13:57
4

I find these types of issues interesting, although it can be hard to answer the question as many different answers are possible depending on how you look at it.

For example, how is driving ability to be measured? Speeding tickets? Accident rate? Surveys of the general driving population?

Perhaps car accidents are a decent indicator given that it is widely accepted that human error is the cause of most accidents. This study states:

According to research done in Germany, human error is the major cause in more than 90 percent of the accidents, and in almost 75 percent of the cases human error is the only cause.

Wikipedia states:

A 1985 report based on British and American crash data found driver error, intoxication and other human factors contribute wholly or partly to about 93% of crashes

I have had a lot of trouble finding reliable statistics on car accidents per manufacturer.

I did find this article (which doesn't show the source of its data) which states that for data collected over 12 months from 2011 to 2011, the top 3 most accident prone cars according to this data were the Honda FR-V, the Volvo XC90 and in third place the Lexus RX.

It is important to note that this in no way indicates that BMW are not the worst drivers (using accidents as a metric) due to the limited amount of data I was able to find. However I could find nothing to support the claim that BMW drivers are the worst, or even any worse than drivers of cars from other manufacturers.

Sonny Ordell
  • 8,695
  • 4
  • 64
  • 102
  • 1
    I don't think that accident rates are even relevant. Some cars are much safer than others, and thus will have less accidents. Some other cars are very difficult to drive, e.g. a Lamborghini, thus the accident rate could be higher. In none of these cases the accident rate measures driver skill. – Sklivvz Jan 21 '12 at 15:20
  • As a side note - I am pretty sure that inexperienced drivers are by far the worst drivers. They pay more insurance for example. – Sklivvz Jan 21 '12 at 15:21
  • 3
    @Sklivvz I considered accident rates a good indicator because by and far, most accidents are caused by human error as opposed to vehicle malfunction. I would have looked at insurance premiums as well but I couldn't find anything for different manufacturers. – Sonny Ordell Jan 21 '12 at 22:17
  • I was thinking about things like ABS etc. which actually reduce the possibility of human error. – Sklivvz Jan 22 '12 at 01:24
  • Even so, according to almost every study/source I could find, human error is responsible for the overwhelming majority of car accidents. – Sonny Ordell Jan 22 '12 at 01:48
  • Ok - so if all accidents were due to human error, safer car brands would have less accidents by preventing human error. If you look at the accidents numbers you are actually looking at which brands are better at preventing human error. I see no reason to believe that bad drivers choose a particular brand. They probably drive slower cars though. – Sklivvz Jan 22 '12 at 01:52
  • Safer car brands can only prevent human error to a limited extent. The fact that there was an accident means the car failed to prevent human error, so it was either a question of driving ability and/or extenuating circumstances. I don't think using accidents alone can show differences in driving ability, however I think it may be useful as an indicator. – Sonny Ordell Jan 22 '12 at 02:53
  • 1
    there's a theory that increasing safety features invites more reckless driving ("people will take the same amount of absolute risk, so making them safer means they can be more reckless without their risk of injury increasing"). Whether that's correct I don't know. – jwenting Feb 08 '12 at 11:47
4

Recently a Dutch tv channel reported on speeding on Dutch roads.The conclusions were based on the number of camera tickets issued between October 2010 and October 2011. They linked speeding to car make. Interestingly, the most speeding drivers are in the age frame 50-59. So if you consider speeding bad driving, there is some "data" (see side node below).

According to their report the drivers of grey mercedes' are the worst followed by Audi and then the BMW.

Some side notes remain. First the journalists who conducted this study only mention absolute numbers without correcting for the distribution of the car models on the Dutch highways. They also only report on their findings without releasing the actual data.

Andra
  • 1,013
  • 1
  • 8
  • 15
  • 1
    and getting speeding tickets doesn't equal irresponsible driving, especially not with automated cameras handing out tickets for 5kmh violations where the speed limit is 120. And you forgot another factor as well: those cars drive a lot more kilometers per year than many others, belonging to people who drive a lot for their jobs (often several hundred kilometers a day, where the average is under 50). What'd be interesting is the number of tickets (or rather accidents) per kilometer driven as a function of the type of car set against the age of the driver. – jwenting Sep 26 '13 at 13:32
  • @jwenting don't forget that if you get a speeding ticket for say 125km/h, your speedometer is probably indicating a bit more then that. – Andra Sep 26 '13 at 14:52
  • it may be, but doesn't have to. Depends on the car and the radar/laser trap used (many of which are deliberately set up to give higher than actual readings). – jwenting Sep 26 '13 at 20:28
0

A recent study suggests that it might not be the model, but the wealth of the driver making him a bad driver. Being an upclass model, could explain why BMW-drivers might drive more inconsiderate then others.

Andra
  • 1,013
  • 1
  • 8
  • 15
  • It's a long bow to draw this conclusion from this experiment. – Oddthinking Sep 26 '13 at 07:24
  • @oddthinking And you think that matching model to driving skills isn't? Behaviour patterns caused by being wealthy, makes a bit more sense then the make of a car model. – Andra Sep 26 '13 at 10:18
  • Or a particular model could encourage poor driving, or have a safety flaw, or a brand could be targeted at nervous drivers, or a model could be attractive to speeders, or... it is easy to speculate. What we ask for here is definitive answers that are based on empirical data, and this answer is neither. – Oddthinking Sep 26 '13 at 11:26
  • @Andra typical slur to "blame the rich" as is so often the case these days... Most accidents are caused by specific age groups, typically either inexperienced, aggressive, young drivers, or experienced drivers who're getting old and haven't yet started to take their increasing reaction times into account. Cars typically driven by those groups then end up on the top of the accident statistics. – jwenting Sep 26 '13 at 13:35
  • exactly - admittedly the fact that we see BMW drivers apparently driving the worst in day to day experience on our roads may be a positive bias in our awareness, or it may be they are just that bad, but how can you measure it...? – Rory Alsop Sep 26 '13 at 13:58
  • 3
    @jwenting do you have data supporting these claims? – Andra Sep 26 '13 at 14:44
  • check the insurance premium tables. It's no surprise those are higher for very young and old drivers... – jwenting Sep 26 '13 at 20:29