Tuesday, January 10, 2012

Freaking out about Walking Drunk

You may have caught the recent Freakonomics broadcast on the perils of walking drunk, a story that's been causing controversy ever since it appeared in the opening chapters of Super-Freakonomics. The quick summary, for those in the dark, is that the Freakonomics authors did some number-crunching and determined that walking drunk is eight times more likely to get you killed than driving drunk.

According to co-author and economist Steven Levitt, "Truly, if you're faced exactly with two choices, walking drunk or driving drunk, you absolutely should drive drunk."

Shockingly, this statement upset some people. But is it true?

We know that in terms of deaths per trip, pedestrians are over-represented when compared to drivers or vehicle passengers. However, even though our Freakonomics friends threw out some assumptions about total miles walked in the US, we don't have great data on overall exposure to walking dangers (e.g. how do you count all those crashes that happen in parking lots?). Thus, contrary to what Levitt and company suggest, it's hard to actually know how dangerous walking is on a per-mile basis.

On the other hand, it's true that walking isn't always safe--particularly if you're walking home after a night at the bars. Nighttime walking, especially on weekends, is highly associated with pedestrian fatalities.  Of course, this is the case whether you're stumbling drunk or stone-cold sober (although it's worth noting that about a third of pedestrian fatalities involve drinking), but I suppose the publicity for your book is better if you focus on perils for tipsy pedestrians...

...which brings us back to Levitt's argument that you should drive, rather than walk, drunk. It could be true that driving is the "correct" choice if you're only considering your personal safety. However, as anyone who's taken Economics 101 could  point out (and has, if you read the comments on the Freakonomics page), there are significant external costs associated with both drunk driving and driving in general. Factor those in, and walking becomes much more appealing from an economic standpoint--but not safer for the pedestrian. Advocates, planners, and traffic engineers: that's your job.


  1. Although I'm just relying on my own faulty memory, it seems that I often hear of drunk-driving crashes where the drunk driver was not harmed at all, but other people were injured or killed. So, perhaps if you're just concerned for your own safety, then driving is better than walking.

    1. ... and driving a tank instead of a car would probably be an even better choice (for you)... :/

  2. Hey, nice blog! I put a link to your blog up on http://streetsblogvermont.blogspot.com/ - hope that's alright.
    Keep up the good work,

  3. Hmm, didn't someone in San Diego take a tank for a joyride once? I'm sure that wasn't safe for pedestrians either...but really, the interesting thing to me is that the Freakinomics folks focused on drunk walking/driving at all--aren't the dangers pedestrians face when sober shocking enough??

    @ streetsblog Vermont: always pleased with publicity, thanks!