In His DUI Platform, Magnuson Shows Correlation Does Imply Causation

September 8th, 2006 at 03:32pm Michael Brylawski 133

In Rick Magnuson's latest post on Pitkin County DUIs, an ad he posted in the Aspen Times (a four-year graph of the fatality rates of Pitkin v. CO), all along with a very balanced article in the Daily News, Magnuson digs into the statistics on Pitkin County DUIs and cites some unpleasant facts:

• Since 2001, there have been 8 alcohol-related vehicle deaths in Pitkin County.
• Given our modest size, this puts our rate at 3x the state average--and 3x that of counties such as Garfield (e.g., Glenwood Springs)
• Pitkin ranks 48th out of the 64 counties in the DUI-related fatality rate

This is not a good record--but statistics on their own are somewhat meaningless. What is often said in doing statistical analysis is that "correlation does not imply causation," i.e., just because statistics indicate a link (DUI deaths and Pitkin County policing), it does not prove it.

In this case, a high death rate--even consistently in the long term as Magnuson showed in his ad--could be due to a host of factors, or even chance. Even though the probability is small, you can flip heads four times in a row. If it was snowing when you did this, you wouldn't assume the snow "caused" you to flip four heads. Or more understandable to this town, maybe Plummer goes four games without throwing an INT, while you were wearing your throwback Elway jersey each week. Was it the jersey? Or like the coin, luck? (trust me, with Plummer, it's luck).

But what is more telling is that Magnuson links a cause to these statistics. He goes beyond the stats to argue that Pitkin County policies and operational strategies may be a causal factor in the DUI rate.

Clearly, Sheriff Braudis is strongly concerned with DUIs and acknowledges them in the Daily News as "statistically the biggest threat to life and limb in Pitkin County." He's not in favor of DUIs (who is, aside from maybe Mel and Paris?). And his Tipsy Taxi program is innovative and progressive. But Magnuson lays out some sobering findings that question whether the Sheriff is doing all that he can do in leading the department to lower the rate:

• Over the last three years (extrapolating 2006), DUI arrests are down 3 fold
• Two out of 19 officers account for 75% of the DUIs
• Pitkin Co. deputies end their shifts at 2am, when the bars close, as opposed to 3am like Aspen, Basalt, and Snowmass
• Only one officer mans the post-2am "graveyard" shift (no pun intended) during weekends
• A common and effective DUI-tool, like police-car cameras, are not in place (anyone whose followed the Gibson case, or seen at least one episode of Cops, can attest to their effectiveness)

Magnuson goes on to propose four initiatives--none of them controversial, all pretty logical--to address these problems, including one that builds on Braudis' most successful program, Tipsy Taxi. He completes his argument, going from DUI stats-->causal factors-->solutions to address the causal factors.

What is dissapointing is Braudis' response. Instead of embracing these suggestions--particularly logical ones like expanding Tipsy Taxi marketing into the bars, in the AM (when people will need to use it), or changing policing shifts--Braudis says "he wouldn't do it any differently."

He cites a demographic shift and the fact that "there are just not as many drunk drivers" as reasons not to change his policy. Both raise questions. On the demographic shift, if people are increasingly moving downvalley, wouldn't that make Pitkin's responsibility greater? It would seem that more people would likely drive drunk downvalley past 2ams as they commute back to the midvalley and Glenwood. Trust me, I love Basalt and El Jebel, but their bar scene ain't Aspen's.

Second, on what statistical basis are there "not as many drunk drivers"? Nationwide stats show DUIs are going up, and Magnuson cites Aspen stats that show the same. Moreover, what is true--and what started this dialogue--is that drunk driving deaths are not going down, and that is the goal.

Frankly, this latest Daily News article, and the previous one where Braudis dismissed Magnuson's green platform (saying "I’d label public safety more important than light bulbs and wind energy,” which now seems to provide unintended irony), shows why I've supported Rick Magnuson from the beginning. Bob Braudis is a good man, but he's been in office, unopposed, for a long time. His office does several things, like Search and Rescue, well. But whether it's getting on board with the Canary Initiative, dealing with sobering DUI stats, confronting the reality of a new-generation of hard drugs (i.e., meth), he really is running on autopilot. He and his office wont change until we, as voters, take action.

I hope people are paying attention to the Sheriff's race the last two weeks, because Magnuson is now hitting issues, analyzing causes, and proposing solutions. He's pushing beyond the tendency of this town to lock into often superficial, yet controversial issues, and start a broader, deeper debate. He's looking and acting not just like a credible candidate--but someone who could be our Sheriff. Even if you are a staunch Braudis supporter, a closely-faught, issues-oriented campaign is a breath of fresh air (or breath with BAC less than 0.08...okay I could not resist).

Entry Filed under: Aspen