You can’t prove a negative.

Prove you are not a liar. It’s an impossible task, you can’t prove that something “didn’t happen” but you can sure get people worked up about it.

Last year a report was published where an anti-gun rights professor created “synthetic” states and compared them to real states that passed right to carry laws, and determined using this method that right to carry laws increase violent crime 10 to 13 percent. When asked about states that had passed right to carry laws and saw no difference, or even a decrease in violent crime rates the author simply stated:

It depends on the state. There was a downward trajectory in crime in the ’90s due to a number of other factors. We are documenting how much a difference it made that states had this law.

Awesome. The author understood that there are “a number of other factors” but made the concrete determination that RTC laws increase crime at a rate of 1% every year (the 10 to 13% at the end of ten years) compared to their “synthetic” state counterpart.

In the words of Mark Twain, “There are lies, damn lies, and statistics.” The report using a synthetic control is much closer to “damn lies” than it is to actual statistical analysis. The first is that without actual context, a percentage is a dimensionless value. Is the violent crime muggings, rape, murder, assault? The authors didn’t bother to share.

In the real world, crime rates generally shift away from crimes against persons (muggings) to crimes against property (burglary) because empty houses don’t shoot back, so the overall crime rate is about the same, with less violent crimes and more other crimes. Most criminals are in it for the money or the thrill, and the ones who are in it for the thrill don’t seem to last long. We have lots of real world data that documents that shift, so clearly the real world data must be wrong if it doesn’t support the “synthetic” data comparison.

John Lott looked deeper into the “synthetic state control” method and found that it took real world data from dissimilar states to create a synthetic state to represent the analog of all the real states that adopted RTC laws. There seems to be no justification for this choice.

Simply comparing California and Texas over a ten year period illustrates that violent crime rates are, 1) centered in large urban areas, 2) exacerbated by trans-national drug cartels, and 3) go up and down regardless of right to carry or “may issue.”  The argument that Texas would be 10% safer than it is without right to carry laws is ludicrous since the violent crime rate tracks so well over the years with California.

Similarly, this is why I don’t like talking about “gun deaths” since counting suicide as a “gun death” doesn’t tell you anything about “violent crime” at all. However anti-gun organizations LOVE to talk about gun deaths because it inflates their numbers, and it is true that having more access to firearms is associated with more suicides by firearms, but NOT a change in the overall suicide rate.

Now John Lott’s data and analysis have been pecked at and analyzed for over two decades now, and so far the most credible rebuke came from the National Research Council which determined that the available data didn’t make a case for an increase or decrease in violent crime based on right to carry laws. In science terms, “we didn’t find a connection between these two things” is a valid outcome.

The biggest problem I have with the “synthetic state” argument is that violent crime rates within states are not uniform, as mentioned earlier they are concentrated in dense urban areas (and sometimes Native American reservations). A state to state comparison is “fuzzy” when it would be more appropriate to look at similar cities. However the US has only a handful of mega-cities, and so comparing Los Angeles to Miami might be valid, but San Francisco to Indianapolis would not. I guess if you took an average of large, medium, and small cities from non-RTC states and factored in socioeconomic and demographic data with the violent crime rates you might be able to make some valid comparisons with large, medium, and small cities from RTC states. That might give some sort of actual insight on the impact of state law on violent crime.

But even then, with good analysis, you still can’t prove a negative.

Here’s a link to Lott’s rebuttal, the “study” is behind a paywall unless you have a .gov email account.

This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to You can’t prove a negative.

  1. Borepatch says:

    You are far too kind to the “researchers” – Eric Raymond has a good overview of Firearms policy fraud committed by “scientists” over the years.

    The errors are all in the direction of supporting more gun control, which supports the use of the word “fraud”. It’s simply not possible that the researchers don’t know about opposing bodies of work (e.g. Lott, Cramer).

    Liked by 1 person

    • rthtgnbs says:

      It doesn’t hurt me to be kind, but honestly the paper should never have passed peer review. But even then, people who should know better make the same mistake, just the other day someone who should know better said that the “Food Pyramid” was probably the biggest killer in American history. She said this without regard for the fact that the average lifespan in the US continually rose from introduction until 2016, and that she couldn’t find an actual control group to compare medical outcomes against. But she had her opinion, and she could have used the “synthetic control” method used in the paper to “prove” her point, but it would be a worthless paper since science is only as rigorous as the control. And when you make an artificial control, you clearly get artificial results.

      Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s