Stand Your Ground laws and homicides

Jeff points me to a paper by Chandler McClellan and Erdal Tekin which begins as follows:

The controversies surrounding Stand Your Ground laws have recently captured the nation’s attention. Since 2005, eighteen states have passed laws extending the right to self-defense with no duty to retreat to any place a person has a legal right to be, and several additional states are debating the adoption of similar legislation. Despite the implications that these laws may have for public safety, there has been little empirical investigation of their impact on crime and victimization. In this paper, we use monthly data from the U.S. Vital Statistics to examine how Stand Your Ground laws affect homicides. We identify the impact of these laws by exploiting variation in the effective date of these laws across states. Our results indicate that Stand Your Ground laws are associated with a significant increase in the number of homicides among whites, especially white males. According to our estimates, between 4.4 and 7.4 additional white males are killed each month as a result of these laws. We find no evidence to suggest that these laws increase homicides among blacks. Our results are robust to a number of specifications and unlikely to be driven entirely by the killings of assailants. Taken together, our findings
raise serious doubts against the argument that Stand Your Ground laws make America safer.

I don’t really trust their regressions. I mean, sure, 5 additional homicides per month is as good an estimate as any, but the conclusions are coming from the data, which McClellan and Tekin display reasonably well:

I think one would want to understand these wavy ups and downs in the curves before making any definitive pronouncements.

From a policy standpoint, I guess it’s no surprise that Stand Your Ground laws could be associated with an increase in homicide. After all, these laws aren’t really enacted as a homicide-control measure, right? It’s more the opposite, that they legalize certain violence that used to be criminal. I could imagine Stand Your Ground decreasing homicide in some sort of deterrence effect, but that would seem to me to be a bit of a bank-shot of an effect, hoping that legalizing some acts of violence would decrease others. It’s possible but I wouldn’t bet on it. To put it another way, even if Stand Your Ground laws really did increase homicides, I could imagine people still supporting the laws on the grounds that some of these homicides were justifiable. I suppose that would be the next stage of research but it would take a lot more effort as it would have to investigate the story of each homicide.

18 thoughts on “Stand Your Ground laws and homicides

  1. According to the CDC, there were 16,799 deaths by violence in 2009. Really hard to get worked up about 60. Apparently, according to one study (http://www.nber.org/papers/w18012), one would expect increasing fuel economy standards by 1 mpg to lead to a larger increase in mortality.

    • Kevin:

      Your comment is reflected in the last sentence of the abstract quoted above: “Taken together, our findings raise serious doubts against the argument that Stand Your Ground laws make America safer.”

      The point is not that Stand Your Ground laws are good or bad but rather that they don’t seem to be making America safer. As I noted in my blog post, this makes sense in that safety doesn’t seem to be the motivation for Stand Your Ground Laws; such laws are about making the “ground” more dangerous. So, indeed, one can support such laws even if they are increasing the number of homicides. (As you indicate, some people are not bothered by the prospect of an additional 60 homicides.)

      • The New York Times reported on April 27, 2005 (about Florida’s SYG), “The bill’s sponsor, Representative Dennis K. Baxley of Ocala, said it would curb violent crime and make citizens feel safer.” That clearly seems like safety is a motivation, prioritizing one’s safety (i.e., the law-abiding citizen) over the criminal. However, I suppose one could feel safer even if one isn’t actually statistically safer.

        http://www.nytimes.com/2005/04/27/national/27shoot.html

      • I’m not quibbling with your post or the paper. Rather, I was making a meta point that a lot of additional investigation and debate over the safety effects of SYG seems like it would be wasted. Even if one thought this study had modest flaws, it would be hard to argue that correcting the flaws would swing the results enough in either direction to justify expending significant resources on it.

        In a rational world, we would just table the safety issue as small beer. Some people might care enough to still argue the moral fairness of SYG (or some other non-safety value of it). Personally, I’d focus my safety-oriented attention on some other topic more likely to produce significant results.

  2. Generally speaking, homicide rates peak in July and bottom in February. (The effect is nontrivial: in 2000-04, average daily homicide rates in July were 28% higher than average rates in February.) So, assuming that there was a strong seasonal pattern in enactment dates of stand-your-ground laws, one could expect to see wavy patterns in figure 2.

    It’s quite common for homicides to occur in the setting where the perpetrator mistakenly believes that his actions are justifiable. Stand-your-ground laws make it more likely for people to think that they are justified, so naturally they increase the number of justified homicides as well as the number of cases where the self-defense claim fails.

    • My thoughts exactly… I suspect deseasonalizing or applying an annual moving average filter prior to alignment and stacking would go a long way in removing the “ups and downs”.

  3. Looking at the “full sample” graph, it looks like there was and is an upward trend in homicides in “Stand Your Ground” states, and a downward trend in the other states. There’s no obvious change in trend at the time where “Stand Your Ground” went into effect. So it looks like it can just be explained by states in which homicides are increasing (for some reason or other) being more likely to enact “Stand Your Ground” laws. That is, the causality may run in the reverse direction from what the authors think.

  4. According to the paper at least some of the models do include month*year fixed effects, although it would be interesting to see the graphs with the month*year effects filtered out — or the predictions of their null models (without SYG effects).

    I wonder if they would be willing/could be convinced to release their data (i.e. the inputs to their regression models)? They presumably can’t release the raw data (the paper says the data are “firearm related homicide victimization between 2000 and
    2009 drawn from the U.S. Vital Statistics. These data are made available from the
    National Center for Health Statistics (NCHS) through a contractual agreement. “), but the month/state aggregated data would be really interesting.

    I’m surprised Andrew didn’t comment on the 6 tables’ worth of summary statistics (including OLS parameters with significance stars) that could have been converted to graphs …

  5. Unless the extra 4-7 weren’t criminals but SYGers who were out of their depth dealing with murderous thugs.

  6. Unlike a rose, a homicide is not a homicide is not a homicide. They should be parsed according to what the FBI calls “circumstance” in its Supplementary Homicide Reports, collected and published annually on the web (Google NACJD). SYG is irrelevant in domestic homicides, for example, and there are other types that are more or less relevant to SYG. I could go into greater depth on this issue, but it’s dinnertime.

  7. Looking at the graphs, I’d like to know what the effect of homicide rate is on whether or not that state passes a stand-your-ground law. The post-pre effect in stand-your-ground states looks like it’s a lot smaller than the difference between SYG and non-SYG states before SYG states actually passed the law.

  8. To put it another way, even if Stand Your Ground laws really did increase homicides, I could imagine people still supporting the laws on the grounds that some of these homicides were justifiable.

    I could imagine that also, and they’d be wrong.

    You (or whoever defends this) are equivocating between two different senses of “justification,” and that is 1) I am within my rights to do X; and 2) X is good. I am within my rights to shoot someone who I believe is attacking me with deadly force, but this is different from saying it is good that they are shot, because that is false. I am within my rights not to put anything in the collection plate, but this is not good.

    • Andy H.:

      I believe that some people think it’s good when civilians shoot criminals. I’m not expressing any opinion on the matter myself; I’m just pointing out that this sort of law can have multiple goals, so someone might support the law even if it were demonstrated to increase the homicide rate.

  9. Pingback: Andrew Gelman on “Stand Your Ground” laws « Andy Hallman's Blog

  10. Pingback: Why U No Llike Control? - Page 16

Comments are closed.