Totovader
Game Warden
- Joined
- Jan 31, 2007
- Messages
- 3,321
I think there's merit to having a discussion about the topics that were raised in recent posts- mainly the massive disparity between what a person carrying a gun means to a proponent of gun control versus a proponent of gun ownership. One looks at that scenario and recoils in fear, the other looks at that scenario and feels relative safety.
I likened the difference to ignorance, bias, and general hoplophobia. Below I will make the case and welcome feedback. In order to be rational, logical folk- we need to be aware of our biases. Reality can help us with this, and that's why skepticism can be so useful: as a methodology it helps us to challenge our concepts, especially those we believe in. Some are content in their beliefs and are not interested in checking them against reality; I simply don't subscribe to that viewpoint and I don't think anyone who calls themselves a skeptic should.
Let's start with an analogy:
If someone has never seen a person of another race before, and all they know of that race is that they are ignorant, violent, etc- all they know is bigotry- is the appropriate response to fear that individual and to cross the street? Or is there an opportunity there to educate ones-self on the facts? What should a rational person do?
Too often, the obvious bigotry comes with this bias: "Don't you watch the news?"
This is an example of availability bias or confirmation bias. If we start out with the belief that Asian people are bad drivers, we will start to see "evidence" of this everywhere. Pretty soon, the "hits" of people who are Asian that cut us off are inflated in our mind and the "misses" of someone who was not Asian who cut us off are minimized in our immediate memory. That person would come home from their daily commute and be more and more convinced every day that Asian's are bad drivers. The bigotry in this example should cue us into questioning these kinds of beliefs- there's nothing about the concept of "Asian" that would make someone inherently a bad driver. Bigots will often chock it up to "culture" but that's just replacing one stereotype with another.
An effective test of this is to include the conditional: "All Asians are bad drivers." When we do that, we can see how important it is to question these kinds of statements.
Maybe you think that all gun owners or all CCW (concealed-carry/carrying a concealed weapon) are violent or potentially violent. Perhaps you watch the news and see how correlated guns are to violence. Perhaps you then look for evidence of this. At the end of the day, you're convinced. When your ideas are challenged, maybe you chock it up to "gun culture" or "gun nuts."
Maybe you think: "Lots of crimes are committed with guns!"
It's true- a lot of crimes are committed with guns. Even though these stats are declining- they're still pretty scary:
source
Those are troubling statistics, indeed. Based on this, your likelihood of being the victim of a crime with a gun is... high? Wait, what is it? Are you likely to be the victim of a gun crime? Perhaps more importantly, does the existence of guns make the crime more likely?
Based on the numbers above, one might conclude that the probability of being murdered by a firearm is 68%, but maybe you've done at least a little bit of math and found that- of the over 10 million crimes per year in the US: 467,321 is an uneasy percentage- about 4.5 percent. I'd be willing to bet that doesn't even matter, though, because that 68% sticks in your head: if we eliminated guns, wouldn't a least a portion of that 68% of murders, 41% of robberies, and 21% of aggravated assaults go away?
And that's where the problems start. This kind of thinking is the result of a very common- and very seductive- error: the confusion of the inverse. When thinking about conditional probabilities (the probability of something given that something else has happened) we can make the mistake of thinking that the probability of A given B is the same as the probability of B given A. It's not.
The probability of a gun being used given that there was a crime is not the same as the probability of a crime given that there is a gun.
Often times that error is used to justify medical arguments against the existence of guns. Medical journals are extremely biased against guns- for a number of different reasons- and they commit this error routinely. When trying to understand how guns affect suicide rates, they will look at suicides, determine how many used guns (65%) and then conclude that the chance of suicide given that there is a gun is 3 times more than those homes without guns- even though the comparison was the chance of a gun, given that there was a suicide. Other studies use the same method to determine the impact on spousal abuse, etc.
With that understanding, what is the actual probability of being the victim of a crime, given that there is a gun? Does the presence of a gun increase your likelihood of being the victime of a crime?
These are hard numbers to get to, to be sure- we have to make some rough estimations and do our best. Using 2011 data: assuming 467,321 crimes had a gun involved, a total of 10.26 million crimes per year, a population of 313 million, and roughly 35% gun ownership here's how it plays out:
The reason your mind has trouble wrapping itself around that is because when you see estimates about victims alone, you aren't taking into account the non-victims when estimating probability. But to truly understand the problem and the probability, we have to account for those things. The number of non-victims is pretty high, and the number of non-victim gun owners is a pretty big variable as well.
Let's be clear: this is not an analysis of whether a gun makes you safer, because I do not distinguish whether the gun in the scenario is owned by the assailant or the victim: I can only go off reported crime and compare that to the number of estimated gun owners (which will contain a population of criminals). To really understand the impact of defensive gun use, we have to look to survey results to get the best possible understanding. All this exercise demonstrates is that the probability we estimate given the available statistics can sometimes lead us to the incorrect estimation of true probability when bias is involved. I am also certainly not saying that the 467,321 victims of crime- some of them homicides- are not important. I am not saying we shouldn't be trying to stop violence or that it isn't a problem. What I am saying is that we can focus our energy on the wrong issues if we allow bias to creep into our thinking. To solve the problem of car accidents, we shouldn't be looking at removing Asian drivers...
Maybe after all this you still think: "I would rather eat in a restaurant where nobody is armed versus a restaurant where everyone is armed!"
Hopefully by now you're starting to even question this statement. But let's pretend you haven't. Let's pretend this bias is still ingrained in your psyche.
Let's say you walk into a convenience store and there are three men standing there: one is wearing camo, has a gun holstered on his hip, and a full beard- he's buying a pack of cigarettes; the second one has a rat tail, a concealed firearm in his back pocket, he's wearing a Gadsden flag t-shirt and just came from the Cliven Bundy ranch- he's buying a disposable cell phone; the third one is a decorated police officer, 10 years on the force and he just saved a little girl from drowning 10 minutes ago- he's buying an anniversary card for his wife.
Who should you be more afraid of?
The error here is the base rate fallacy: taking into account factors which are irrelevant and excluding the prior probabilities. This is central to the problem of gun bias.
Pretty simple statistics are involved, here. Again, there are assumptions to be sure- but I took the best possible numbers. There were 247 fatalities from police officers involved in misconduct in 2010- this would be unjustified homicides, not accidental bystander stuff. That same year, CCW permit holders committed 67 murders. Remember: we can't just compare those two numbers and say you are 3.7 times more likely to die by a corrupt cop than by a CCW permit holder, we have to take the non-corrupt cops and non-murderous permit holders into account- otherwise we're just saying that if you died, what are the chances it was from one of the two. There are way more permit holders than cops, so it makes the disparity even bigger.
The rate of murder by corrupt officers is 32 per 100k, and for CCW permit holders it's .7 per 100k. That means: as an innocent person you are 46 times more likely to die by a corrupt cop than by a CCW permit holder.
I'm not saying don't trust cops: 32 per 100k is a 1 in 3097 chance that a cop is a killer, about the same chance of being hit by a meteorite. The point here is that comparitively, CCW permit holders as a group are more law-abiding, less violent, and as a result safer than the people we trust implicitly. It's an odd phenomenon to be sure, because it's hard to look at those three people and say that you should be more worried about the cop, but the statistics just don't corroborate the fear of CCW permit holders.
Think of that next time you happen to be out and see someone with a gun in their holster. For the same reason that you don't think twice when it's a cop or a soldier, perhaps you should try to eliminate some of your bias and examine the facts.
I likened the difference to ignorance, bias, and general hoplophobia. Below I will make the case and welcome feedback. In order to be rational, logical folk- we need to be aware of our biases. Reality can help us with this, and that's why skepticism can be so useful: as a methodology it helps us to challenge our concepts, especially those we believe in. Some are content in their beliefs and are not interested in checking them against reality; I simply don't subscribe to that viewpoint and I don't think anyone who calls themselves a skeptic should.
Let's start with an analogy:
If someone has never seen a person of another race before, and all they know of that race is that they are ignorant, violent, etc- all they know is bigotry- is the appropriate response to fear that individual and to cross the street? Or is there an opportunity there to educate ones-self on the facts? What should a rational person do?
Too often, the obvious bigotry comes with this bias: "Don't you watch the news?"
This is an example of availability bias or confirmation bias. If we start out with the belief that Asian people are bad drivers, we will start to see "evidence" of this everywhere. Pretty soon, the "hits" of people who are Asian that cut us off are inflated in our mind and the "misses" of someone who was not Asian who cut us off are minimized in our immediate memory. That person would come home from their daily commute and be more and more convinced every day that Asian's are bad drivers. The bigotry in this example should cue us into questioning these kinds of beliefs- there's nothing about the concept of "Asian" that would make someone inherently a bad driver. Bigots will often chock it up to "culture" but that's just replacing one stereotype with another.
An effective test of this is to include the conditional: "All Asians are bad drivers." When we do that, we can see how important it is to question these kinds of statements.
Maybe you think that all gun owners or all CCW (concealed-carry/carrying a concealed weapon) are violent or potentially violent. Perhaps you watch the news and see how correlated guns are to violence. Perhaps you then look for evidence of this. At the end of the day, you're convinced. When your ideas are challenged, maybe you chock it up to "gun culture" or "gun nuts."
Maybe you think: "Lots of crimes are committed with guns!"
It's true- a lot of crimes are committed with guns. Even though these stats are declining- they're still pretty scary:
source
Those are troubling statistics, indeed. Based on this, your likelihood of being the victim of a crime with a gun is... high? Wait, what is it? Are you likely to be the victim of a gun crime? Perhaps more importantly, does the existence of guns make the crime more likely?
Based on the numbers above, one might conclude that the probability of being murdered by a firearm is 68%, but maybe you've done at least a little bit of math and found that- of the over 10 million crimes per year in the US: 467,321 is an uneasy percentage- about 4.5 percent. I'd be willing to bet that doesn't even matter, though, because that 68% sticks in your head: if we eliminated guns, wouldn't a least a portion of that 68% of murders, 41% of robberies, and 21% of aggravated assaults go away?
And that's where the problems start. This kind of thinking is the result of a very common- and very seductive- error: the confusion of the inverse. When thinking about conditional probabilities (the probability of something given that something else has happened) we can make the mistake of thinking that the probability of A given B is the same as the probability of B given A. It's not.
The probability of a gun being used given that there was a crime is not the same as the probability of a crime given that there is a gun.
Often times that error is used to justify medical arguments against the existence of guns. Medical journals are extremely biased against guns- for a number of different reasons- and they commit this error routinely. When trying to understand how guns affect suicide rates, they will look at suicides, determine how many used guns (65%) and then conclude that the chance of suicide given that there is a gun is 3 times more than those homes without guns- even though the comparison was the chance of a gun, given that there was a suicide. Other studies use the same method to determine the impact on spousal abuse, etc.
With that understanding, what is the actual probability of being the victim of a crime, given that there is a gun? Does the presence of a gun increase your likelihood of being the victime of a crime?
These are hard numbers to get to, to be sure- we have to make some rough estimations and do our best. Using 2011 data: assuming 467,321 crimes had a gun involved, a total of 10.26 million crimes per year, a population of 313 million, and roughly 35% gun ownership here's how it plays out:
- Probability of a firearm given that you have been the victim of a crime = 4.55% (~1 in 22 chance) - an interesting statistic, but not what we're after
- Probability that you are a victim of crime given that there is a firearm = .43% (~1 in 234)
- Probability that you are the victim of a crime given that there is no firearm = 4.81% (~1 in 21)
- That shakes out to: you are 11 times more likely to be the victim of a crime given there is no firearm
The reason your mind has trouble wrapping itself around that is because when you see estimates about victims alone, you aren't taking into account the non-victims when estimating probability. But to truly understand the problem and the probability, we have to account for those things. The number of non-victims is pretty high, and the number of non-victim gun owners is a pretty big variable as well.
Let's be clear: this is not an analysis of whether a gun makes you safer, because I do not distinguish whether the gun in the scenario is owned by the assailant or the victim: I can only go off reported crime and compare that to the number of estimated gun owners (which will contain a population of criminals). To really understand the impact of defensive gun use, we have to look to survey results to get the best possible understanding. All this exercise demonstrates is that the probability we estimate given the available statistics can sometimes lead us to the incorrect estimation of true probability when bias is involved. I am also certainly not saying that the 467,321 victims of crime- some of them homicides- are not important. I am not saying we shouldn't be trying to stop violence or that it isn't a problem. What I am saying is that we can focus our energy on the wrong issues if we allow bias to creep into our thinking. To solve the problem of car accidents, we shouldn't be looking at removing Asian drivers...
Maybe after all this you still think: "I would rather eat in a restaurant where nobody is armed versus a restaurant where everyone is armed!"
Hopefully by now you're starting to even question this statement. But let's pretend you haven't. Let's pretend this bias is still ingrained in your psyche.
Let's say you walk into a convenience store and there are three men standing there: one is wearing camo, has a gun holstered on his hip, and a full beard- he's buying a pack of cigarettes; the second one has a rat tail, a concealed firearm in his back pocket, he's wearing a Gadsden flag t-shirt and just came from the Cliven Bundy ranch- he's buying a disposable cell phone; the third one is a decorated police officer, 10 years on the force and he just saved a little girl from drowning 10 minutes ago- he's buying an anniversary card for his wife.
Who should you be more afraid of?
- Probably the camo guy. Who wears camo unless they are up to something? Plus: the beard. Never trust people with beards.
- Maybe the concealed carry guy? The people at the Bundy ranch are nuts. He's clearly up to something.
- Police, though? Clearly this is a set-up, but how could it possibly be the police officer? Police are trained, they are sworn to duty, they prevent crime... why should I be afraid of a police officer?
The error here is the base rate fallacy: taking into account factors which are irrelevant and excluding the prior probabilities. This is central to the problem of gun bias.
Pretty simple statistics are involved, here. Again, there are assumptions to be sure- but I took the best possible numbers. There were 247 fatalities from police officers involved in misconduct in 2010- this would be unjustified homicides, not accidental bystander stuff. That same year, CCW permit holders committed 67 murders. Remember: we can't just compare those two numbers and say you are 3.7 times more likely to die by a corrupt cop than by a CCW permit holder, we have to take the non-corrupt cops and non-murderous permit holders into account- otherwise we're just saying that if you died, what are the chances it was from one of the two. There are way more permit holders than cops, so it makes the disparity even bigger.
The rate of murder by corrupt officers is 32 per 100k, and for CCW permit holders it's .7 per 100k. That means: as an innocent person you are 46 times more likely to die by a corrupt cop than by a CCW permit holder.
I'm not saying don't trust cops: 32 per 100k is a 1 in 3097 chance that a cop is a killer, about the same chance of being hit by a meteorite. The point here is that comparitively, CCW permit holders as a group are more law-abiding, less violent, and as a result safer than the people we trust implicitly. It's an odd phenomenon to be sure, because it's hard to look at those three people and say that you should be more worried about the cop, but the statistics just don't corroborate the fear of CCW permit holders.
Think of that next time you happen to be out and see someone with a gun in their holster. For the same reason that you don't think twice when it's a cop or a soldier, perhaps you should try to eliminate some of your bias and examine the facts.
