The U.S. is Not as Violent as You Think

Since the shooting in Newtown there has been hundreds of articles written about guns, gun control, and gun violence. Patrick Cahalan has written the best of the bunch:

Basically, you’re not very likely to be murdered in either [the U.S. or the U.K.


In fact, given that 10% of the murders in the U.S. are family-related, and given that 74% of murder defendants have a prior criminal record, and 44% of murder victims have a prior criminal record, and that 82% of non-family murder victims are male, and 53% of non-family murder victims are black, and 38.5% of non-family murder victims are between 20-29, and that half of murders involve alcohol… the actual odds that you’ll be the victim of a murder, if you’re a sober, law-abiding middle aged white person who doesn’t piss off your wife or your children… is actually probably pretty close to the general murder rate in the U.K.


How to Write a Love Letter and Salvage Valentine’s Day (A Guide for Men)
Day Jail for Kids
Teenagers Spend Too Much Time Around Teenagers
Where Are You Most Likely to Run Over a Deer?
  • Damien

    But the same is true in the UK as well. It doesn’t make much sense to compare the best-case scenario in the US to the UK average. I understand that the odds of being murdered are actually quite small in either case. But that second paragraph confuses the issues unnecessarily and, at worst, it might be interpreted as trivializing the deaths of people who are not like “us” (should I feel more comfortable that it’s really inner-city black youths who are the most likely to be murdered?). Of course if you take out of your sample all the people who are actually more likely to be murdered, you discover that the remaining people are very unlikely to be murdered. Why the fact that they’re still as likely to be murdered as an unbiased sample of UK people is cause for celebration is beyond me.

    In fact the whole paragraph at the source site is fairly mistaken. People should not denounce statistical manipulation unless they can actually get their own figures right. 3.2 per 100,000 is NOT 0.32% but 0.0032%. Likewise, it’s 0.0001 and not 0.01%. You can get the rate for an individual by dividing by 100.000 but the reason it’s not done is not manipulation but because you’re unlikely to be murdered in any given year and the per-person figures are very hard to size up. It doesn’t change anything to the fact that the relative risk of being murdered in the US is 32 times higher.

    • DZ

      Wow, awesome deconstruction of a really poorly considered post.

  • Patrick Cahalan


    Yes, the same is true in the UK as well.

    The point is that the UK and the US are not directly comparable, and comparing “the US” to “other high income nations” or “other Western European Nations” is bad statistical analysis; comparing an outlier with an aggregated data set instead of comparing the outlier to the individual nations. You see this all the time in public policy debates and beginning social science research classes will tell you this is a big no-no.

    Moreover, most people who worry about “the violence in America” aren’t aware of how unlikely it is that they, in particular, will be victims of violence, thus they look at a scary number and incorrectly conflate it with their own risk profile.

    The U.S. has dissimilarities with the UK that make it a poor direct comparison. The UK does not have a War on Drugs the way we do (a likely major contributing factor) nor do they have pernicious generational poverty for large groups of minorities.

    As for the decimal error, I was writing a blog post, not a research paper. Thanks for catching that, I’ll fix it.

    “the reason it’s not done is not manipulation but because you’re unlikely to be murdered in any given year and the per-person figures are very hard to size up.”

    If you’re sufficiently unlikely to be murdered in a given year, a comparison to another country where you are also unlikely to be murdered in a given year is not warranted. Because you’re unlikely to be murdered in a given year.

    The fact that you’re – generally – 32 times more likely to be murdered in the US instead of the UK is a sound bite. That’s all it is.

  • Patrick Cahalan

    Oh, and thanks for the compliment, Joe. I think that might be the first time you and I agreed on anything :)