Skip to content
 

“Texting bans don’t reduce crashes; effects are slight crash increases”

John Christie sends along this. As someone who owns neither a car nor a mobile phone, it’s hard for me to relate to this one, but it’s certainly a classic example for teaching causal inference.

7 Comments

  1. dearl says:

    I understand not having a car if you live or work in a major city, but no cell phone? What's the weather like back there in 1995? ;)

  2. Andrew Gelman says:

    As in 1995, I can go about my daily life without people expecting that they can reach me at all times.

  3. Joseph says:

    I have an iPhone for the internet capability but I often leave the ringer off. So I almost never actually recieve a call on it and life seems to work out pretty well.

  4. Kaiser says:

    Great teaching material. Make sure you read to the bottom of the paper when it mentions the most important issue: noncompliance. As a matter of communication, I don't like the fact that this word was not mentioned in the first paragraph. If noncompliance is an issue, then all the prior analysis is just junk.

  5. Musal says:

    I agree with the comment above…
    "Noncompliance is a likely reason texting bans aren't reducing crashes. Survey results indicate that many drivers, especially younger ones, shrug off these bans. "
    As an avid runner and biker I take extreme offense at such junk lobbyist research (I presume it is lobbyist because the claim is so preposterous) which seems to imply texting while driving does not cause increases in crashes. If the main point was about noncompliance that should be either part of the research or the emphasis. Not the statement at the bottom of the page.

    How many times I saved myself in the last second from being hit because the driver of the SUV was texting in his/her truck (no an SUV is not a car) I lost count.

    Yes I actually got hit by a driver that was texting while making a left turn into the incoming traffic.

  6. Ben Bolker says:

    Why does noncompliance make the conclusions
    junk?

    If the assumptions of the analysis are correct then
    what the study is telling us is that *texting bans
    are not effective*, not that *texting does not
    increase crashes*. The policy response should presumably
    be in figuring out ways of devising more effective
    regulation (increased fines, education, fascist
    technical solutions like blocking texts from moving
    vehicles …)

    The article actually seems reasonably precise
    in stating its conclusions.

  7. Kaiser says:

    Ben: I know we are splitting hairs. But say I do a clinical trial of drug A vs. drug B. Drug A is a pill; drug B requires a painful injection. I analyze the results and conclude that drug A is much more effective than drug B. In the footnote, I let on that the people assigned to drug B just refuse to take the injections. You can say well drug B is ineffective and it is ineffective because no one would take it. But if I state drug A is better than drug B, most readers would think drug B doesn't work because it does not cure the disease, it's not typical that readers would think people won't take the drug.