Here's an interesting statistic for you: http://www.news.com.au/couriermail/story/0,23739,25321352-3102,00.html

"out of 52,128 random breath tests, 225 people were caught drink-driving."

IOW, less than 0.5% of those tested were at or beyond the (already arguably too low) 0.5% limit. The sample is quite large, so this would seem to be a reasonably robust result.

Here's another statistic: "In the 48 hours to midnight Friday, three motorists were caught drug-driving,"

A whole 3 in the entire State of Qld. There was not a road fatality in Qld for the period reported on.

Given the hoopla around the issue of driving while impaired, with all sorts of broad claims made about the "scourge of drugs" and the massive expenditure by the various state govts on kits to test for them, surely only 3 people caught in a 48hour public holiday blitz is unexpected?

In NSW, http://www.smh.com.au/national/increase-in-road-deaths-alarms-police-20090411-a3du.html the figures are even lower, with "less than 0.2 per cent of the 93,500 drivers randomly breath tested over the weekend so far found to be over the limit." In NSW, the number of motoring deaths has risen by 20% over last year.

What is wrong? Is the testing methodology unreliable? Are police not using it, or using it incorrectly? Could it be that the so-called hordes of druggies on the road are simply a figment of political and police imagination? I suspect the latter.

My questions for OLOers sre these: does the small number of people using alcohol or drugs and then driving continue to justify the massive expenditure and the assault on civil liberties that roadside testing implies?

Is alcohol/drug use actually a causative factor in a large number of accidents, or is it merely a convenient scapegoat that can be used to give police more intrusive powers? Even the police only claim 2 out of 19 pedestrian deaths over the period "may" have alcohol as a "significant" factor.

the last time i perused the govt figures[1999]the figures had been dumbed down,the figures were representitive of thousands[and alternativly scaled down to per 1000 units

its clever accounting[in one case it was reported 950/1000 positive drink driving results, [they of course forgot to manipulate the figures down..lol]

but the same thing re fed figures[re drug deaths ,by reason, smokong was ATTRIBUTED to have 19,000 deaths [out of 22,000]..lol, mainly because docters ATTRIBUTED it [smoking] as THE CAUSE OF DEATH[lol death from smoking is pure fiction[but then if docters write a 'CAUSE' of death there is no autopsy[yeah its legal muder]..but the ant-smoking lobby needs its stats

asbestosis ans all cancer deaths [lung diseases are ALL attrributable to smoking, thus no asbestosis claims thanks be to docters atribut-ing death due to smoking 19,000 deaths from the act of smoking needs to be contrasted to the deaths from booze..lol

there were 3,000 of them[but in the final numbers that was somehow [via some aqccounting trick modified down to 1800[by reading the notes it was because of the benifitial affects of a drink...lol

so 1200 of 3000 who died from booze didnt..lol

anyhow the same tricks apply thoughout the whole scam of statistics, read them they are played down or up as suits the current debait

like full time employment means working 2 or more hours [lol]
real inflation is running a lot higher but we removed housing and a few other things from the equation

here is a fact hidden from people[deaths by adverse reaction to PERSCRIBED DRUGS]..in usa 3 jumbo jets [equivelent] die EACH DAY,in usa along, thats adverse reaction to perscribed by a docter drugs

then we habe the dose needed to cure number[i forget its acronim], but it lists the number needing to take a medication to get ONE CURE, like heart medications need 60 to take the medicine TO GET ONE CURE, the 60 have no advantage [and most likely adverse reaction, [1 in 10 hospital admitions is adverse reaction[1 in 10 result in death]

yeah give us the REAL FIGURES

"My questions for OLOers sre these: does the small number of people using alcohol or drugs and then driving continue to justify the massive expenditure and the assault on civil liberties that roadside testing implies?"

You would have to compare them with the figures before roadside testing? If roadside testing has reduced the number to a much safer level then perhaps it is justified. I don't see roadside testing as an affront of civil liberties - I think in my entire life I have only ever been pulled over four times.

I know from growing up in the era before RBT that there were many more people getting into cars while intoxiacted and unable to exercise sound judgement. There were more accidents before random breath testing and before the 0.5 limit was introduced.

Statistically the number of road accidents and road fatalities has declined despite growing populations so that is a good thing.

It is not surprising that fewer people are caught drink-driving over holiday periods when it is known that police and RBT are more prevalent.

"There's Lies, Damn Lies, and Statistics".
But on a more serious level, hey, my kids are out there driving on those roads, so go for it, test away, and I don't mind the delay when I get pulled over for one.
I'm a social drinker, and I simply don't drive if I am drinking, end of, so I've never had the least worry blowing in the bag. Anyone can choose to do the same, if they don't, get them boys!

What's a life or six worth?
It only takes one irresponsible drug infused person to kill or maim .
Two cases come to mind
The woman in Surfers who while on drugs did a burn out ,lost it and ploughed into pedestrians killing two injuring 4.
Then 6 youths in a ute killing four 3 girls in the back.
Stats form no part in my reasoning on road safety I am concerned about people not gambling with lives.

Has it occurred to anyone that maybe random drug/alcohol testing is having a deterrent effect and that is why there are so few arrests?

To take current statistics without also looking at stats from before RBT is obviously misleading and the worst possible spin imaginable.

A far more useful investigation along the same lines would be to compare statistics before and after 911 and whether all the fences, guards and screening everywhere has been worth it