Polls v. facts

My Facebook page consists of links to all my blog posts automatically published by WordPress. A couple of friends from many decades back (it’s been 40 years since I’ve seen Ken) questioned my unsubstantiated claim that the cartoonist in this blog post was delusional to believe gun ownership rates were going up instead of down.

It is true that various polls give us the idea there is a slight downward trend. The following data is from Gallup:



The cartoon in question was published in 2013. It could have been in response to Gallup’s annual poll.

As with all statistics there are errors. In many cases increasing the sample size will reduce the errors. In other cases changing your sample methods will decrease the errors. Many initial studies are done by grad students on fellow classmates in their college or university. The sample bias is obvious and many poll questions would be pointless in such an environment. There is also a danger of the respondent giving erroneous data and in some cases would also make many poll questions pointless.

For some polls there is little or no danger of getting erroneous data from the respondents. Asking someone their favorite color would be quite safe from this type of error. Asking if there is a wanted fugitive justice living in their home would be at extremely high risk from this type of error with a resultant bias toward “No” approaching 100%. Asking them if they are in an active homosexual relationship would probably have a mild bias in the “No” direction today but would have had much higher bias in the “No” direction 75 or 100 years ago.

It is the hypothesis of many people, myself included, that gun ownership polls suffers from a minor to moderate bias in the “No” direction today.

If you look carefully at the data above you will find something I think is very telling. Look at the rate of people that have “No opinion” in the tables above. I suspect this predominately represents those people that do own guns or have guns in their homes and do not wish to lie to the pollster. Notice how this was consistently 1% in the time frame 2000 to 2008 but came up after Obama was elected? A similar pattern shows up after his reelection in 2012. It is my hypothesis that most or perhaps all of this change is due to concerns that polls about guns could be used to build databases of gun owners for nefarious purposes by an administration hostile to gun ownership. This hypothesis is further supported by the data in the first table going back to the Clinton administration. It is also supported by public approval ratings of the NRA being in the 50+% range while government approval ratings are in the 10-% range.

I’ve made a case the “No opinion” data above reflects gun owners who own guns but aren’t willing to lie to the pollster. Now lets look at indications gun owners are lying and pushing the sample bias even further in the “No” direction.

It is well known that FBI NICS checks show dramatic increases in gun sales in recent years and only deviated from the upward trend slightly in 2014 (original here). The claim by anti-gun people is that the polling data demonstrates existing gun owners are purchasing more guns and that there are few new gun owners. Those of us close to the gun community strongly suspect this is false because of what we experience at gun stores, the ranges, and in our personal lives. The impression is there are a lot of new gun owners. But how can we quantify that and avoid our own confirmation, and other, biases?

One way to quantify it is through data from states that require registration of gun owners. Illinois is one such state. A Firearm Owner Identification (FOID) card is required for gun ownership. Here is what the State of Illinois says on their website about FOID cards:

The Illinois State Police (ISP), Firearms Services Bureau, (FSB) has received a record number of FOID card applications since May 2012. For example, in January 2013, the ISP FSB, received 61,172 FOID applications. As a comparison, in January 2012, the ISP FSB, received 31,655, which had been the highest number of FOID applications received during the month of January in years prior to 2012.

Those are applications. It is unclear but it appears these applications do not include renewals. If true then January 2013 had nearly double the number of new gun owners apply that it had in January 2012. This cannot be attributed to growth in the state population:

Illinois is one of the slowest growing states in the US. Between 2000 and 2010, the population of Illinois grew by a mere 3.3%, compared to a national average US population growth of 9.7%.

Estimates for 2010 through 2014 indicate even lower growth rates.

There are also sources with more detailed information about FOID numbers. Even if FOID stats includes renewals it is conclusive evidence that the rate of legal gun ownership in Illinois is increasing, not decreasing.

Massachusetts has a similar Firearm Identification (FID) card but I was unable to find the numbers (Weer’d Beard gives us these which are consistent even if not complete). Ditto for California, New Jersey and New York (I stand corrected). If someone has those numbers please let me know and I’ll update this post with the additional data.

Another proxy for gun ownership rates is the concealed weapon permits. DiveMedic gives us some numbers that also strongly suggest an increase in gun ownership rates:

… the number of concealed weapons permits in the state of Florida has more than doubled since 2009, when there were 591,830, which was double the number of permits from 2002, when there were 295,658 permit holders.

My conclusion is the polls on firearm ownership rates have a significant negative bias that distorts the true rates and that the true rate is much higher than it was 10 years ago. This is supported by the polls themselves, hard data from registration of gun owners, and proxy data. I have not been able to find any data with contradicts this conclusion.


25 thoughts on “Polls v. facts

    • I had seen both of Kevin’s post before I started writing mine. But I wanted to address gun ownership rates not absolute numbers. Kevin’s posts primarily address absolute numbers and I didn’t want to muddy the focus of my post.

  1. Is it also possible that even if the percentage of homes is going down, that the number of homes is going up?
    30% of 100 homes= 30 homes. 20% of 500 homes= 100 homes.

  2. I looked over that entire quite long page. With one small example, there is no clue what the sample size is. (The one exception is the 2nd amendment question, which says it was asked of 373 gun owners and 630 non-owners.) So there is no way to tell whether the margin of error is 1% or 20%. My guess is that the rather small fluctuations reflect sampling errors; there is no real trend visible here. The fact Gallup doesn’t list the sample size means they are incompetent (if we want to give them the benefit of the doubt).
    John Lott, in “More guns, less crime” and/or Gary Kleck in his study of defensive gun use talked at great length about sampling biases, explaining the limitations of polls that ask about guns because of people’s concern about disclosing such information to strangers. Those issues clearly apply here as well. The “no opinion” column presumably actually means “refused to answer” and reflects some of that number. I expect that a substantial fraction of the “no” answer is from people who do own guns but don’t want to say so to random callers. (Conversely, I would expect the “yes” answers to be almost 100% truthful, since there is no obvious reason for a non-owner to claim to be a gun owner.)
    It’s interesting that the 2nd amendment question says that 6% of gun owners believe the fiction that the 2nd amendment protects state militias. It’s also interesting that of non-owners, 28% believe that; it’s encouraging that the number is as small as that given the widespread propaganda for that lie.
    Some other curious/amusing/interesting answers: 50% believe that courtrooms would be less safe if judges were armed. And 73% believe that schools would be more dangerous if “school officials” were armed. I guess I can understand that if “officials” meant principals, who as a group seem to be mentally unqualified for just about anything you could imagine. But still, that’s rather an indictment of the public’s opinion of school personnel…

  3. Ring ring ring
    Me: Hello?
    Caller: Do You..
    Me: click.

    I probably have contributed to that negative bias.

    • With all due respect, you’re doing it wrong.

      [Ring ring ring…]
      Me: Hello?
      Caller: Do you…
      Me: hold on a second [puts phone down, wanders off to continue what I was doing]

      It’s great to give a no-response to phone pollsters, but doing it my way ties them up longer.

  4. I went into some detail about the Florida CCW question in today’s post over at my blog.

    I also want to add, in response to the question about school personnel:
    This is my first year as a high school biology teacher after I retired from 23 years as a firefighter and paramedic, where I spent part of my career working with the SWAT team. I spent years as an IDPA competitor, and I am a military veteran. I have carried a concealed weapon for more than 25 years.

    I have spent more than 30 years in one uniform or another, defending the people who could not defend themselves, saving lives, and helping others. In that time, I have had dozens of background checks for security clearances, teaching, firefighting, and paramedic licenses, as well as for concealed weapons permits. Again and again over the past three decades, I have proven my character, my devotion, and my trustworthiness.

    I would, if necessary, lay down my life in defense of the children that have been placed in my care. Even in Kindergarten. Possibly YOUR children, if you are reading this.

    Except the politicians have declared that I am not permitted to do so, because they don’t trust me. So instead, I must sit in the dark and wait for help that may not come, wait with your children to die at the hands of a madman who didn’t obey your laws or your signs.

    • Oh, hey, a bio class? So, you had done a little ecology field trip, and showed them what sort of living stuff you can dig up in a single scoop of swamp muck in that nice, properly sharpened shovel you still haven’t gotten around to taking home yet, right? Right?
      I love science classes – all sorts of good reasons to have multipurpose tools around.
      Oh, and welcome to the profession.

      • I took a biology class in college. We went on a field trip. At the first stop, someone upset a swarm of bees and we had to run the entire route. What was supposed to be a three hour walk ended up taking about 15 minutes.

  5. Some Massachusetts numbers here

    Also just doing a news search for carry permit applications and issuance isn’t as clean as ownership permits, but still while states like Florida or Virginia, or North Carolina there is a benefit to having a permit even if you don’t carry, there is NO benefit to having a permit if you don’t own a gun.

    Massive upticks in carry permit applications, as well as NICS check, as well as the recent run on firearms, and the current drought on .22 LR could all be indicative of other factors, but together only point to ownership going up, and going up in a non-trivial amount.

  6. If ownership is going down, why did my fish and game club have to institute a membership cap for the first time in its 50 year history?

    And why are so many of those new member first time gun owners?

    And why am I seeing this across multiple gun clubs in my neck of the woods?

    The plural of anecdote is not data, but I see a trend which indicates the opposite of the poll data.

    None of this is news of course to the regulars here, and I have little doubt we are all seeing the same thing.


  7. The uptick in Florida permits may very well indicate a nationwide trend, not just state trend. Since the Florida CCW is coveted as one of a handful that are not only granted to non-residents, but is accepted in turn by a large number of states, the data very likely indicates new permittees from home states applying next for the Florida one.

  8. Minor quibbles: New York state does not have an FID card. Long guns are not registered, and there is no statewide permitting system to own guns. What we have is a permitting system to own/carry hand guns. To buy a rifle/shotgun in NYS, all you need to do is pass NICS. In my experience, using pistol permit numbers as a proxy for numbers of gun owners may be problematic. It’s about 18 months from the time you drop a completed application off at the county building (permitting is done county by county with some being de facto shall issue and others severely restricting carry permits) and most people I know that get permits were already gun owners. People will likely not wait 18 months to get their first gun when they can get a long gun same day.

    • Corrected. Thanks.

      I was thinking they had long gun registration which would indirectly give us gun owner registration.

  9. I’ve worked part time behind the counter at one of the largest gunstore’s in southern NJ since 2003, when I started we had a three year old pistol range and we had one NRA certified pistol instructor on staff. He rarely had classes and normally sold guns, ammo ect. Today we have six pistol instructors on staff, and only two have ever worked sales. Each schedules his own classes with those interested in classes, and someone is having class almost every evening. Almost all students are new gun owners or have just applied for NJ’s FID card. I’m familiar with anecdotes =/= data but something has definitely changed.

    • Since 2003, have other gun stores closed? Has anyone who used to give classes quit giving classes? Does your store offer better customer service, etc. than other nearby stores?

      There are a lot of factors that can cause an increase in business. Obviously, your store is doing something right.

      • One new gunstore in the last five years, also giving classes within 20 miles, I was unaware of anyone else in the area giving lessons ten years ago.

    • The numbers are from a poll performed in the late 1990’s. I’m not seeing how this relates to claims in the recent polls that gun ownership rates are decreasing. Am I missing something?

    • You know…. thinking about your rate/percentage data….

      If Los Angeles County had an ownership rate of 14%, other places would have to increase a huge amount to overcome that. There are 10 million people in Los Angeles County. The whole state of Idaho has less than 2 million. The entire state of Washington is around 7 million.

      I used to know a lot of gun owners when I lived in Florida. Around here, I don’t know hardly any. (I’ve met two gun collectors, but one moved to Nevada.) I don’t know anyone who goes target shooting or anyone who hunts.

      I wonder what the gun ownership rate is in New York City? In Washington DC?

  10. Pingback: Tuesday News Links 03-10-2015 | Shall Not Be Questioned

Comments are closed.