Author Topic: Problem with the Levitt "Guns vs. Pools" statistic  (Read 7661 times)

0 Members and 1 Guest are viewing this topic.

Offline CZattorneyFL

  • Jr. Member
  • **
  • Posts: 61
Problem with the Levitt "Guns vs. Pools" statistic
« on: November 21, 2005, 03:19:39 PM »
Hey all. Lots of people, including myself, have been quoting "Freakonomics" by Steven Levitt for the proposition that having a gun in your home is 100 times safer than having a pool in your backyard. But I've been thinking about that statistic a lot, and, unfortunately, it doesn't hold up.

Levitt starts with 4 facts:

1. There are about 6 million residential pools in the US.

2. There are an estimated 200 million guns in the US.

3. About 550 kids under 10 drown each year in residential swimming pools.

4. About 175 kids under 10 die each year from accidents involving guns.

The conclusion Levitt wants you to draw (and that I wanted to draw) is obvious: if you have kids, having guns in your house is statistically much safer than having a pool in your backyard. But you can't draw that conclusion from these facts without more information. The statistic suffers from the same disease that ruins many statistics: comparing apples to oranges.

For starters, he's comparing "residential pools" to "guns". We probably know that those pools are in houses, but we have no idea where those guns are. Just for the sake of argument, what if that gun count included, say, US military small arms? What if that count included illegal weapons? Gang weapons? If the gun count is over-inclusive, and doesn't really count just "residential guns" there'd be a lot fewer "residential guns" per child death, and having a gun in your house wouldn't be so much safer than having a pool.

But there are even more questions:

What if people tend to own more than one gun, but not more than one pool? Then you'd have a lot fewer "gun houses" per child death as well.

Also, even if Levitt is talking purely about guns legally owned in homes, we don't know whether the percentage of homes with guns AND kids is anywhere near the percentage of homes with pools AND kids. Without that, we can't say whether owning a gun in your home is safer for your kids than having a pool in the backyard.

The point is that we don't know enough about the guns Levitt is counting to make the point he wants to make. Now, obviously, I'm not telling you whether guns or pools are safer for the kids in your house; I'm just saying we don't know from Levitt's cute little comparison. And I'd be tickled pink if someone made a more thorough comparison and factored in the instances where having a gun in the house actually SAVED a kid's life!

Anyway, this is especially annoying because I was excited to quote someone who was interested in the gun safety issue from a purely intellectual standpoint. Too bad Steven Levitt didn't use his intellect quite enough!

Here's a link to some professor who quotes the Levitt book directly and (I think) basically makes the same point I'm making.

profesora.blogspot.com/20...ished.html

Offline dvsnse

  • Full Member
  • ***
  • Posts: 180
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #1 on: November 21, 2005, 04:51:14 PM »
Good point about taking statistics at face value.  What did Mark Twain say? Something like "there are lies, bleep lies, and statistics."  

But are the two essential number correct?  If so, then the fact remains that over three times more children die each year from swimming pool accidents than from accidental shootings.  
I think the actual number of dead kids is far more telling than the probabilities; especially when you start talking about serious injuries and deaths involving automobiles or bicycles.

Offline CZattorneyFL

  • Jr. Member
  • **
  • Posts: 61
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #2 on: November 21, 2005, 05:01:29 PM »
Stephen,

Initially, I was inclined to come to the same conclusion as you: even if we chop the gun count down drastically, and there are only as many houses with guns and kids as there are houses with pools and kids, pools are still about three times as dangerous as guns in the house.  This still shows that guns are a lot safer than other objects that people aren't insanely afraid of keeping in their houses, and it reveals that a lot of peoples' fear of owning guns stems from emotions, not reason.

Unfortunately, I still wouldn't place too much weight on that argument.  Sometimes I've found statistics, realized they had problems, swapped out the faulty assumptions for more realistic ones (or even drastically conservative ones), and I've still been way off.  At the end of the day, we need to either get a more honest and precise study, or we should just keep relying on our brains.  Either way, my bet is that the Mossberg stays in the bedroom ;)

BRASMAN

  • Guest
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #3 on: November 22, 2005, 04:13:38 AM »
I would say that there are a lot more guns that are in houses than pools and if that is so if more kids are dying from pool accidents than gun accidents I would still say that guns are safer than pools.

Offline CZattorneyFL

  • Jr. Member
  • **
  • Posts: 61
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #4 on: November 22, 2005, 08:40:32 AM »
Brasman,

Yeah, I'm inclined to agree with you, but with the same reservations as above.  It's unlikely that 200 million would be knocked down to less than 6 million... Maybe the NRA has some harder information on the number of houses with guns, maybe split up by age? Then we could make some more informed estimations...

BRASMAN

  • Guest
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #5 on: November 22, 2005, 12:33:13 PM »
I guess what I am saying is I think they could only prove it MORE and that it is proved.

Offline CZattorneyFL

  • Jr. Member
  • **
  • Posts: 61
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #6 on: November 22, 2005, 01:28:35 PM »
Brasman,

I'm not sure I understood what you were saying.

Levitt is comparing apples to oranges: he's comparing pools in homes to guns, without saying whether those guns are in homes or not.  If the guns he's counting are not in homes, the comparison isn't valid: we won't know whether having a gun in the home is more or less dangerous than having a pool in the home.  

Now, probably, not all of the 200 million guns are in houses.  So that number (200 million) can't go up, it can only go down (if, say, some of the guns Levitt is counting aren't in houses, like the pools are).  If the "200 million guns" number goes down, that makes child-deaths-per-gun more common,  making guns more dangerous.

Right?

BRASMAN

  • Guest
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #7 on: November 23, 2005, 02:56:53 AM »
What I am saying is I KNOW that there are more guns in homes than pools. Are all the pools in the homes of criminals who will drown people or responsible adults or adults that are not responsible? Are there more drowning deaths than firearm accident deaths period. A good friend of mines son who was my youngest sons best friend shot himself on accident while on a camping outing. I consider that in the home because the pistol came from the home and it was a houshold event. You could throw in all kinds of variables. The fact is though either more kids died drowning in pools or firearms accidents. I will say pools are more dangerous period in the home or not. However my point was I know that more homes have firearms in them than homes have pools so the firearms will certainly out number the pools.

Offline CZattorneyFL

  • Jr. Member
  • **
  • Posts: 61
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #8 on: November 23, 2005, 08:15:17 AM »
I agree with you that, statistically speaking, guns in the home are probably safer than pools in the home.  But I don't understand your certainty.

How do you know that there are more guns in homes than pools in homes?  Ever been to South Florida?;)

BRASMAN

  • Guest
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #9 on: November 23, 2005, 03:11:40 PM »
Yes I have been to south florida. Ecinomics and appartments. I know a few people around the country that have pools but most everyone I know including most of the antigun people I have met have guns in their house. I do not need a scientific study when I can look around and see it for my self. I have heard "Trust us we are the experts" to many times.

Offline Jeff Bergquist

  • Hero Member
  • *****
  • Posts: 602
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #10 on: November 25, 2005, 12:02:16 AM »
Quote
Quote:
1. There are about 6 million residential pools in the US.

2. There are an estimated 200 million guns in the US.

3. About 550 kids under 10 drown each year in residential swimming pools.

4. About 175 kids under 10 die each year from accidents involving guns.


I understand your hesitation to compare such possibly unrelated statistics, the fact that might link them together better is that there are guns in about half the homes in america, which if I remember correctly means guns in some 40 million homes. (I'm operating from memory so someone might want to verify my numbers but the point remains valid.) Now compare that to 6 million homes with pools to get a more viable ratio. The numbers I come up with is one drowning death of a child under 10 for each 11,000 houses with pools, and 1 firearm death of a child under 10 for each 229,000 homes with guns. So it can reasonably be argued that having a pool is over 20 times more dangerous to children under 10 than having one or more guns, all else being equal.

Jeff
The bold type giveth, the fine print taketh away.

Offline EZ CZ75

  • Hero Member
  • *****
  • Posts: 363
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #11 on: November 25, 2005, 04:02:30 AM »
Jeff, the latest "stats" say that there are about 60 million homes with firearms at this point.  I am sure that number is from either legal gun registrations or polls.

I believe that it doesn't really matter which is more deadly.  Yes, it is nice ammo for the anti-gun enthusiasts, but it still means that there were more than 100 accidental child gun deaths last year.  I believe that education is the only way to ever drop that number.  Eliminating the number would be nice, but not realistic.

My own recent research has pointed me in the direction that the states that are more open with guns and education tend to have less accidents.  That is hard to prove, but an educated child is more likely to survive a gun-related incident than is a sheltered child.  I believe this so strongly that even my two-year-old daughter knows that she can't play with guns and only gets to shoot with an adult.

Knowledge is power.

BRASMAN

  • Guest
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #12 on: November 25, 2005, 04:34:25 AM »
Ditto.

Offline CZattorneyFL

  • Jr. Member
  • **
  • Posts: 61
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #13 on: November 25, 2005, 09:26:05 AM »
Hey, that's great news if it's true; we might actually end up with a significantly meaningful statistic if we can arrive at the actual number of homes with guns.

Where did you guys get that 40-60 million houses with guns number?

Offline Jeff Bergquist

  • Hero Member
  • *****
  • Posts: 602
Problem with the Levitt "Guns vs. Pools" statistic
« Reply #14 on: November 25, 2005, 11:44:55 AM »
I believe my numbers were based on survey results and statistics that were being disseminated by the NRA a while back; possibly DOJ or FBI statistics; you might check their websites.

Jeff
The bold type giveth, the fine print taketh away.