Previous Post
Next Post

“Americans who have been recent crime victims report higher rates of gun ownership than those who have not been victims,” the pollsters at Gallup report. “Thirty-three percent of U.S. adults who have been recent victims of assault, theft or property crimes own a gun, compared with 28% of those who have not been victims — a statistically significant difference.” Or is it?

Although it is not possible to know from the survey questions whether the crime prompted the individual to buy a gun or if the person owned a gun before the crime occurred, the modest yet significant relationship between recent crime victimization and gun ownership is clear.

In other words, the Gallup poll doesn’t tell us squat. While we can guess that being victimized leads people to tool-up, it could also be true that being a gun owner makes a person a more likely target for criminals. For reasons we could only wonder about — because Gallup failed to ask any of the pertinent questions. A fact they admit, again, in their conclusion.

Although the analysis demonstrates a statistically significant relationship between crime victimization and gun ownership, it cannot answer why the relationship exists. An obvious explanation is some of those who have been a crime victim purchase a gun as a reaction to that event. The Gallup data do not explore when the gun purchase was made in relation to when the crime occurred, so it is not possible to know to what extent this explains the relationship.

So what, exactly, was the point of this exercise? Hell if I know.

U.S. gun policy has come under increased scrutiny in recent decades, driven partly by mass shootings but also by the high rate of gun-related homicides. For crime victims, the threat of victimization is no longer a possibility but a reality. Crime victims’ desire to protect themselves may explain why many gun owners do not favor stricter gun laws, and why gun owners as well as nonowners are reluctant to back outright bans on guns.

People who’ve been the victims of crime know that they could be a victim of a crime? Who knew? Other than everybody. Note to Gallup: must try harder.

Previous Post
Next Post


  1. Wonder how many of the crime victims own a car. That comparison would be relating crime experience to ownership of a different object.

    But, maybe the poll tells us more people own guns than the media want to know about.

  2. Yup, that is just about the most pointless poll ever. Also, since Gallup usually presents an error bar of plus-minus 3-4%, I’m wondering how 5% is really “statistically significant. “

  3. This is the stupidest question ever. If I have a lawn ornament stolen out of my front yard while at work I have been victim of property crime. It has virtually zero relevance to firearms ownership.

    It could be simple as a gun owner is less likely to be urban, and therefore the crime rates are lower.

    • True, but people whose possessions are stolen, are more likely to have possessions stolen because they have them.

      Be interesting to see the results of a poll among those who had possessions to steal, and those who didn’t.

      Got a feeling that having possessions is a major factor in the number of thefts recorded.

  4. Ummm…I don’t tell Gallup(or anyone else) if I’m a gun owner. Except HERE …pathetic “poll”.

  5. In other news, Gallup found that 96% of victims of vehicle theft also owned vehicles.

    Gallup also found that people who did not own a bike experience a 99.9% reduction in their chance of having their bicycle stolen.

    Gallup also found that 87% of crime victims had eaten on the day of their victimization.

  6. Morons need a lesson on how to interpret statistics (accurately). Specifically what you can and can not infer from measurement of a conditional probability.

    The probability that you are a crime victim in the past 12 months given that you’ve had an interaction with police in the past 12 months is probably higher than the probability of the general population being a crime victim in the past 12 months. Does that mean police cause crime victims?

  7. So of the gun owners they asked 33% were victims of crime and 28% weren’t? What about the other 39%, were they sorta victims but it didn’t really count?

    • The main question was in reference to crime victims, not gun ownership. And those two numbers aren’t part of the same 100% total.

      Of crime victims, 33% are gun owners (67% are not).
      Of non-victims, 28% are gun owners (72% are not).

  8. Gallup regularly introduces questions that they haven’t yet fully fleshed out. It’s part of their ongoing R&D. Essentially they’re testing the question out and then looking for changes in the answer rates/patterns as they modify the question and related questions.

    It’s how they build a better set of questions to get better and more reliable data. If you’re on their panel (someone they question regularly with a standard set of questions to monitor changes in attitudes) they generally identify the test questions for you when they ask them. Sometimes they don’t but the questions are easy to spot when you’ve been a panelist for awhile.

    This question appeared, in slightly different form, maybe six months ago as a subset of questions for regular panelists. I guess now they’re trying to compare that data to general answer data.

    • Accepting the explanation, the level of intelligence in Gallup’s R&D group is not confidence-building. Maybe they could save time and money simply asking, “What do you think about xxxxxx?

      • Not really because everyone’s answer would vary to some degree and that sort of defeats the purpose of polling by making the answers difficult to categorize. The expense would rise and the extraction of useful data would be impossible.

        Think of it this way: TTAG asks about a specific gun which, by some strange occurrence, we all happen to own or have in our possession for some period of time.

        Do you like the gun, why or why not?

        That question has a nearly infinite number of answers. Some people like it. Some people hate it. Some people are lukewarm. Some people are in between those three options. Some people like/don’t like the trigger/sights/thickness/grip/breaking weight/ergos/caliber/takedown procedure/availability of parts in their area/brand/amount of writing on the slide/whatever. Other people like/don’t like the aesthetics for a ton of different reasons. Maybe this gun doesn’t digest their favorite JHP as well as it might etc etc. Of course you’ll also have some guy who doesn’t like it because it reminds him of his ex or his favorite dog that was run over or whatever.

        The list becomes nearly endless and difficult if not impossible to break down to numbers that mean anything when you sample a large number of people. For TTAG they might get 200 comments. Gallup will get thousands per month.

        That’s why you don’t get open ended questions in polling and when you do get an “open ended” question you get a selection of answers from which the respondent must pick one or more that “describe your feelings the best”. Presidential polls often contain “other” and don’t mention who that other is because they’re not going to bother noting that 0.000001% of people responded “Mickey Mouse” while 0.000002% preferred Daffy Duck.

        • So, the polling they are reporting is more scientific, useful, enlightening and statistically valid than just asking a completely open-ended, generalized question, and categorizing the result?

          If they are going to publish useless information, why go to all the trouble to pretend they are serious players in the polling business?

  9. I am going to go out on a limb and suggest:
    (a) some areas have quite a bit more crime than other areas
    (b) a relatively high percentage of people who live in those relatively high crime areas recognize the utility of having a firearm and acquire one
    (c) a relatively high percentage of people who live in relatively low crime areas do not see a compelling reason to acquire a firearm and therefore do not acquire a firearm

    And I was able to figure that out in spite of the fact that I never studied journalism or political science.

    • Way over-thinking it. Put simply (by someone else already), people who xxxx have a greater probability of yyyyyy, than people who do not xxxx.

      • “I have property valuable enough to be at higher risk of being stolen, therefore I am more likely to be a victim of the crime of theft than someone without sh*t worth stealing. I also am of sufficient means and stability to be on a panel regularly question by Gallup.”

        Sorry for the straw man, but regardless, their pool kinda seems self-selecting for certain traits.

    • I would suggest the exact opposite is true. Those in low crime (rural and suburbia) are vastly more likely to own guns than those in high crime (heavily urbanized) areas.

      In fact, ever poll on the subject bears this out.

  10. Have you been to the doctor for something other than a routine checkup in the last 12 months?
    Do you own firearms?

    Results: Some people who own guns got sick this year and some people who own guns didn’t.

  11. So how long until the media twists this to claim that those who do not own guns are less likely to be a victim of a crime?

  12. “Gallup asks Americans whether in the past 12 months they personally have been the victim of a number of crimes, including burglary, property theft, assault and vandalism.”

    So, if I own a gun and away from my home when someone breasks and enters, steals things or vandalizes my home, how is my possesion of a gun while I am away going to prevent that?

    Its a meaningless survey.

    Why not conclude that my neighbor’s alarm system doesn’t do a thing at MY home?
    Silly Gallup…

Comments are closed.