Correlation is not causation

Reputo demonstrates that if someone claims the firearm homicide reductions in the late 90s were influenced by the anti-gun laws passed in ’93 and ’94 then those same laws can be claimed to have reduced the homicides by sharp objects and suffocation.

Nice job!

Share

5 thoughts on “Correlation is not causation

  1. I have heard the “correlation does not equal causation” until I am sick of it. Here’s why. Those mopes who say that leave out a very important aspect. While the statement in and of itself is true, the lie of ommission is committed when it is not also stated that “causation does equal correlation”.

    Until we are willing to look at the entire proposition I am going to assume that so many correlations in all venues of “more guns, less crime” at the very least, indicates there must be some causation. Else the correlation would not be so ubiquitous across so many socio-economic, geographical, and cultural venues.

    Address that first. The proper strategy or tactic if you prefer, is to secure our persons and neighborhoods first. We have to place fear of retaliation in the minds of the criminal to get him to stop robbing, killing, raping before we can address the issues the liberals blame for crime rampant in only certain sections of our society. Any fool can see this. Hell, Helen Keller could see that.
    There have been several studies now conducted of convicted violent offenders and in each it was revealed that they feared running up against an armed citizen far more than they feared the police. That fear of a possibly armed prospective victim cannot help but be a causative factor in the correlation of more guns in the hands of citizens being found in venues of lowering violent crime rates. Especially when that lowered violent crime rate has been so consistent across all venues where citizens are armed.

  2. Correct straightarrow, “causation does equal correlation.” Where people misunderstand is that all causes correlate with their outcomes. Not all things correlated with an outcome are causes (in some cases the outcomes are a direct result of the same cause). To think of it another way, a rectangle is defined as a four sided shape with 4 right angles. A square fits this definition. But all rectangles are not squares. All squares are rectangles.

    The reason why you hear correlation/causation ad nauseum is because that vast majority of people don’t understand the two terms and use them interchangeably. They are not interchangeable.

  3. then those same laws can be claimed to have reduced the homicides by sharp objects and suffocation.

    Believe it or not, there was one left-wing poster (on the defunct Intellectual Capital site in the late Nineties) who actually believed that. His theory was that guns encouraged all violence, including non-gun violence.

    Ironically, this same guy was always challenging opposing posters to fights (and then backing out when they told him either that he’d have to foot the airline bill, or else that they saw no reasons to limit themselves to fists against an attacker :)).

  4. Leaving a comment at Reputo’s site seems to be too “difficult” (I don’t have or desire to get any of the required accounts) so I’ll leave it here:
    From the linked article:
    “The fact that all three curves are tracking so well (and in some cases have nearly identical year over year changes) means that these three causes of homicide are highly correlated.”

    “… these three causes …” should be “… these three methods …” or something similar. The _cause_ of a homicide would require information not provided by rate and count data. The use of the word “cause” in this way would seem to imply just what many of the gun grabbers claim: mere possession of a gun _is_ the cause of many homicides and that if the guns were not available then the homicides would not occur.

  5. Thanks for the catch Bob R. I checked the post and I use “method” in other instances. I corrected it. I write them off of the top of my head and rarely go back to double check spelling, grammar, and punctuation.

Comments are closed.