Tuesday, December 20, 2005

Hello, Occifer.

Hello, Occifer. What's the problem? - Hiccup-

Came across this article by Max Borders via tcsdaily about the problems with using blood alcohol standards in policing our roads. Readers might want to take a peek.

Part of the problem with analyzing and talking about these type of issues is the moral issues that many attach to these things. People that have a moral aversion to drinking bring a certain amout of bigotry to this issue.

For example, if I'm in my car at 11:00 PM on a Friday and I'm hit by another car, and neither me or the driver of the other vehicle have been drinking, it's just an ordinary accident.

Take the same time and situation and say the other driver had had a few beers, and instead of it being an ordinary accident, I become the victim of a "drunk driver". A whole new dimension is added here because this other fellow quaffed a beaker or two of brew a while earlier. Now, assuming the other driver drank enough beers to substantially impair his driving, it would be correct to say that I was a victim of a drunk driver. However, what if the beers consumed that evening didn't impair the other driver? How should he be treated before the law?

It's certainly true that very high levels of blood alcohol content certainly impair one's ability to operate a vehicle safely. But for the vast majority of weekend revelers, they don't reach those high levels of BAC.

I always like to do the math on these types of things. Are there drinking related accidents somewhere every weekend? Sure there are. But how many people are out and about on Fridays and Saturdays making merry with alcohol? How many millions of miles do these revelers drive every weekend? Across the nation, millions, perhaps tens of millions of people are out at bars, night clubs, parties, fraternal organizations, sports venues, etc., making merry with alcohol. And you know what? About 99.9999999% of them make it home that evening and live to make merry another day.

If alcohol consumption makes drivers impaired, you'd think we'd have thousands, perhaps tens of thousands of accidents each weekend. But we don't.

Part of the issue here is that people don't understand risks and risk analysis in these type of things. For example (pulling numbers from thin air here for example purposes only), let's say that anytime I get in my car and pull onto the street, the odds of me having an accident on that trip are 2000 to 1 against me having an accident. That doesn't mean that I will have an accident when trip 2000 rolls around, just that the odds of me having an accident are very low for EVERY trip.

Now, let's say that I've had a few glasses of wine before pulling out on the street. My odds of having an accident have increased, but the question is by how much have the odds worsened. If I've only had a couple of glasses, the odds of me having an accident might be 1950 to 1 against me having an accident. Have the odds worsened? The answer is likely "yes", but not by much. Now, if I've finished off 2 bottles of wine before entering the street, the odds of me having a problem on the road will likely be a lot worse, say 400 to 1 that I will not have an accident. Have the odds worsened? Yes? But the odds are still in my favor, but not nearly as much.

For too many people, the fact that some drinkers do have accidents means that all drinkers will have accidents sooner or later. But that's not the case. I might drink 2 bottles of wine on a thousand separate occassions, then drive afterwords and never get in an accident. In other words, odds are just that, odds. It doesn't mean a particular outcome will happen.

Another example, flipping a coin. The odds are 50:50, right? Yet you may do 1,000 flips and record 600 heads and 400 tails. Does that mean the odds are 60:40? No, the odds are always 50:50 for each flip of the coin. You might do another 1,000 flips, and the total for the 2,000 flips might be 975 heads and 1,025 tails, but the odds for each INDIVIDUAL flip are still 50:50, heads or tails.

Now, getting back to DUI's. As Radley Balko notes on his blog, most serious accidents involving DUI's show that the drunk driver has a BAC of 0.14 or higher. Does that mean that every driver with a BAC of .14 or higher will have an accident? No, it only means they are more likely to have an accident as compared to those drivers that have a BAC lower than 0.14.

I guess my point is this. A tragedy is still a tragedy, whether the driver that caused the tragedy is at 0.14 BAC, 0.12 BAC, 0.06 BAC, or 0.0 BAC. The aversion some of us have to alcohol colors our assessement of these things. You combine that with a misundstanding of odds and risk and all you get is just another example of moralizing, and not analysis.

This is type of thinking is very similar to the thinking people have that support drug prohibitions. They reason that because some users DO become addicts, it is inevitable that ALL users will become addicts, therefore no one should use drugs and prohibition is warranted. That's just not the case folks.


Post a Comment

<< Home