Recently, HuffPo, WaPo, et. al. Presented this report (which I link here for you to download) The Impact of Right to Carry Laws and the NRC
Report: The Latest Lessons for the Empirical
Evaluation of Law and Policy to ‘prove’ more guns == more crime, I will include selected quotes from it as I go on to debunk it here.
Statistics or Non-Stats?
From the Abstract of the report:
Across the basic seven Index I crime categories, the strongest evidence of a statistically significant effect would be
for aggravated assault, with 11 of 28 estimates suggesting that RTC laws increase this crime at the .10 confidence
level. An omitted variable bias test on our preferred Table 8a results suggests that our estimated 8 percent increase
in aggravated assaults from RTC laws may understate the true harmful impact of RTC laws on aggravated assault,
which may explain why this finding is only significant at the .10 level in many of our models. Our analysis of the
year-by-year impact of RTC laws also suggests that RTC laws increase aggravated assaults. Our analysis of
admittedly imperfect gun aggravated assaults provides suggestive evidence that RTC laws may be associated with
large increases in this crime, perhaps increasing such gun assaults by almost 33 percent.
So this is accurate to a .10 confidence level? Well let us see what the definition of a Confidence Level is. From here: A confidence level is defined as this :
A confidence level refers to the percentage of all possible samples that can be expected to include the true population parameter. For example, suppose all possible samples were selected from the same population, and a confidence interval were computed for each sample. A 95% confidence level implies that 95% of the confidence intervals would include the true population parameter.
Now either they don’t understand the terms in common usage among statisticians, or they are telling us something very important here. They are saying there is only a 10% chance that their report is accurate. If the true confidence level is .10, as defined above in common usage it means that the true parameter used is only 10% likely to fall into their numbers.
I will of course leave it open to the idea that they simply misused the term confidence level, or mean something different from the common usage.
Models or Reality?
Again from the abstract.
In addition to aggravated assault, the most plausible state models conducted over the entire 1979-2010 period
provide evidence that RTC laws increase rape and robbery (but usually only at the .10 level). In contrast, for the
period from 1999-2010 (which seeks to remove the confounding influence of the crack cocaine epidemic), the
preferred state model (for those who accept the Wolfers proposition that one should not control for state trends)
yields statistically significant evidence for only one crime — suggesting that RTC laws increase the rate of murder at
the .05 significance level.
Since 1993, the year of peak murder in the US, murder has dropped significantly across the nation. In addition RTC is now in more states.
1994: Alaska, Arizona, Tennessee, and Wyoming; 1995: Arkansas, Nevada*, North Carolina, Oklahoma, Texas, Utah,* and Virginia*; 1996: Kentucky, Louisiana,* and South Carolina*; 2001: Michigan*; 2003: Colorado*; New Mexico, Minnesota,* and Missouri; 2004: Ohio; 2006: Kansas and Nebraska; 2010: Iowa,* and 2011: Wisconsin.
Since 1995 is the earliest Year that the current UCR has, I will have to count Crime numbers from now
(using 2011, though I will include 2012 numbers I ended up using 2013 numbers ) until back to the Year before RTC laws were passed in each state.
State Total : Murder Number (Rate) ; Violent Crime Number (rate)
Starting with Kentucky, Louisiana, and South Carolina
1995 (FBI UCR)
State total : Murder 276 (7.2) ; Violent crime 14,079 (364.7)
State total: Murder 167 (3.8) ; Violent crime 8,737 (198.8)
1995 (FBI UCR)
State Total : Murder 740 (17.0) ; Violent crime 43,741 (1007.4)
State Total : Murder 498 (10.8) ; Violent crime 23,609 (510.4)
1995 (FBI UCR)
State Total : Murder 292 (7.9) ; Violent crime 36,067 (981.9)
State Total : Murder 297 (6.2) ; Violent Crime 23,625 (494.8)
2000 (FBI UCR)
State Total : Murder 669 (6.7); Violent crime 55,159 (555.0)
State Total : Murder 631 (6.4); Violent crime 42,536 (429.8)
2002 (FBI UCR)
State Total: Murder 179 (4.0); Violent Crime 15,882 (352.4)
State Total: Murder 178 (3.4)*; Violent Crime 15,342 (291.2)
2002 (FBI UCR)
State Total: Murder 152 (8.2); Violent Crime 13,719 (739.5)
State Total: Murder 125 (6.0); Violent Crime 16,496 (591.2)
2002 (FBI UCR)
State Total: Murder 112 (2.2); Violent Crime 13,428 (267.5)
State Total: Murder 114 (2.1)*; Violent Crime 12,100 (223.2)
2002 (FBI UCR)
State Total: Murder 331 (5.8); Violent Crime 30,557(538.7)
State Total: Murder 371 (6.1); Violent Crime 25,509 (422)
State Total: Murder 522 (4.6) ; Violent Crime 38,103 (333.2)
State Total: Murder 455 (3.9) ; Violent Crime 31,904 (275.7)
State Total: Murder 102 (3.7) ; Violent Crime 10,634 (387.4)
State Total: Murder 112 (3.9)* ; Violent Crime 9,478 (327.5)
State Total: Murder 44 (2.5) ; Violent Crime 5,048 (287.0)
State Total: Murder 57(3.1) ; Violent Crime 4,712 (252.2)
State Total: Murder 34 (1.1) ; Violent Crime 8,397 (279.2)
State Total: Murder 43 (1.4) ; Violent Crime 8,062 (260.9)
State Total: Murder 155 (2.7) ; Violent Crime 14,142 (248.7)
State Total: Murder 162 (2.8) ; Violent Crime 15,570 (271.1)
Of the 13 states that have added RTC laws since 1995, the earliest point for which data is accessible online via FBI UCR, 5 states have an increase in murders, and only one has an increase in total violent crime. Of these 5 states none has a murder rate that has increased by more than .5 per 100,000 people. Of the remaining 8 states, the average decrease in murder rate was 1.9 per 100,000. The overall average for the 13 states sampled was a drop in the murder rate of 1 per 100,000.
Violent crime overall, except in one state, Wisconsin, has gone down across the board.
* : States marked with * had an unusual increase in the year over year murders for the year marked, this increase is more than 14% from the previous year, and may be an indication of some other factor. Despite this variation, these are the numbers used to maintain consistency, use of 2012 numbers would have made more of a positive effect on the outcome of this quick comparison in the favor of the pro-gun argument.
These are statewide numbers, and the opposition used county numbers. Will there be a significant difference if we use the county numbers? No, as even if one county increased, the rest of the state would have decreased to compensate.
So why did they use county numbers? My answer is that they didn’t check each and every county, there are 3000+ counties in the US, by taking a ‘representative’ sample of counties they could create a model to do the work for them. This is what they did in fact, despite my opinion that 3000 counties isn’t a significant hurdle to doing accurate research.
Let us Introduce ourselves to reality.
John Lott and David Mustard initiated the “More Guns, Less Crime” discussion with their widely cited 1997 paper arguing that the adoption of RTC laws has played a major role in reducing violent crime. However, as Ayres and Donohue (2003a) note, Lott and Mustard’s period of analysis ended just before the extraordinary crime drop of the 1990s. They concluded that extending Lott and Mustard’s dataset beyond 1992 undermined the “More Guns, Less Crime” (MGLC) hypothesis.
The problem here, is that prior to 1989 few states had any RTC laws at all. As I showed above 13 states didn’t do RTC until after the period discussed here. 42 States are currently RTC states, with only 8 states not being such. Crime is down, and yes the big crime drop started occurring in the 1990’s as suggested above, but as mentioned before in 1986 there were few states that had RTC. In fact the big crime drop in the 90’s comes at the same time as RTC was being picked up by new states, with only 16 being such in 1993 and a large group joining in 1994. It was here that crime started dropping, unlike what is implied by the above paragraph the drop was anything but extraordinary. Dropping only a few percent per year each year, but dropping steadily, 4.1% drop in 1995 as compared to 1994, a drop of 7% from 1996 to 1995, another 4% per year in 1997 over 1996. (FBI UCR)The ‘extraordinary’ drops coinciding with increased access to firearms each year. Not claiming causation here, but it certainly doesn’t show what the authors of this report indicate.
The author goes on to present some pretty graphs on page 10-14, Charts comparing murder, rape, etc. rates vs. year spread out among ‘early, middle, late, and non RTC’ states. While interesting, they don’t show causation, it is possible that the crime rates in these states are the reasons they adopted RTC when they did. More to the point, this treats states such as CT, where I live, as not an RTC state, as it is technically a ‘may-issue’ state. CT, like some other ‘may-issue’ states are essentially ‘shall-issue’ in practice.
… analysis of the county data set from 1977-1997 using the Lott-Mustard specification (revised to measure state-specific effects) indicated that RTC laws across all states raised total crime costs by as much as $524 million.
Increasing the costs of crime, includes things like increased policing, prosecution, and medical costs resulting from DGU, as well as investigation costs when a DGU occurs, in addition I don’t see any mention of inflation adjusting the pricing here.
Crime is a complex issue, and gun control is as well. It is true that crime dropped in the mid to late 1990’s, was the fact that more states became RTC a part of this? Crack didn’t just disappear in the mid-90’s and in fact is just as readily out there, the idea that somehow the crack wars simply ended, seems off the mark. Likely what happened is an increase in police presence, a general improvement in the economy thanks to the dot-com bubble, and RTC all played a part in reducing crime. Using the approved time period from the study, we would find that of the 13 states that became RTC states, most saw crime drop, and the crime drop was significant even compared to the crime increase of the few that saw an increase. In addition the usage of RTC as a metric is confusing, considering that three states, CT, DE, RI have some level of ‘shall issue’ even if the state is ‘may issue’ on the books. Also, long term impact is easily as important. The point being that the data, simply doesn’t support more guns == more crime, and it honestly only gives little, insignificant support to more guns == less crime.
As I always like to say, 2A isn’t about any of this anyway, this would be covered by 10A which states:
The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.
Since this was a right reserved by the people, it is still a right reserved by the people.