More Stand-Your-Ground, More Murder
June 14, 2012 11:29 AM   Subscribe

In addition to removing the duty to retreat when outside the home, Florida's 2005 "Stand Your Ground" law removed the civil liability to offenders who had acted within the law and added a presumption of reasonable belief of imminent harm necessitating a lethal response. These three elements were present in over 20 other state laws similar to Florida's. The following NBER working paper by two Texas A&M economists provides new statistical evidence that these laws caused a 7 to 9 percent increase in homicides and non-negligent manslaughter. Consider this post a companion to this previously, as well as this previously.

Abstract:

"Since Florida adopted the first castle doctrine law in 2005, more than 20 other states have passed similar self-defense laws that justify the use of deadly force in a wider set of circumstances. Elements of these laws include removing the duty to retreat in places outside of one’s home, adding a presumption of reasonable belief of imminent harm necessitating a lethal response, and removing civil liability for those acting under the law. This paper examines whether aiding self-defense in this way deters crime or, alternatively, escalates violence. To do so, we apply a difference-in-differences research design by exploiting the within-state variation in law adoption. We find no evidence of deterrence; burglary, robbery, and aggravated assault are unaffected by the laws. On the other hand, we find that murder and non-negligent manslaughter are increased by 7 to 9 percent. This could represent either increased use of lethal force in self-defense situations, or the escalation of violence in otherwise non-lethal situations. Regardless, the results indicate that a primary consequence of strengthening self-defense law is increased homicide."

Media coverage at Business Insider, WSJ, Media Matters, and Dallas Observer.

For those interested in the statistical methodology used, it's called a form of multivariate regression analysis commonly called "differences in differences". Imbens and Wooldridge NBER lecture notes are useful (slides). Emmanuel Saez's lecture notes are also nice if you want to compare. The theory of causality that diff-in-diff (or DiD, or DD) ascribes to is often called the Rubin-Neyman causal model, or sometimes just the potential outcomes model where causality is measured as the difference between the "actual outcome" and the "potential outcome" for a given unit with and without treatment, respectively.
posted by scunning (40 comments total) 14 users marked this as a favorite
 
the results indicate that a primary consequence of strengthening self-defense law is increased homicide.

Now that's what I call circular reasoning. Since homicide can be defined as "the deliberate and unlawful killing of one person by another"--without any regard for the act's legality, a law that strengthens self-defense will inevitably increase homicide. Try again, docs.
posted by Ardiril at 11:36 AM on June 14, 2012 [2 favorites]


What? Your objection makes no sense. "The deliberate and unlawful killing of one person by another" makes explicit reference to the act's legality. Do you have a substantive critique that is not incorrect?
posted by OmieWise at 11:39 AM on June 14, 2012 [2 favorites]


Sorry, should have included emphasis: ""the deliberate and unlawful killing of one person by another"
posted by OmieWise at 11:39 AM on June 14, 2012


Read the whole thing Ardiril before you say they did a bad job. They find no effect on property crime, but they do find large increases in homicides. So whatever kind of homicide increase this is, it does not appear to be blocking any actual property crimes. If they are defending themselves and killing the perpetrator who would have otherwise committed a property crime, wouldn't we expect to find a reduction in property crimes in addition to an increase in homicides by your reasoning? But they don't find that. No effect on property crimes at all. And given hierarchy rule which suppresses the lesser crime and reports only the worse crime when more than one offense occurs in a single incident, you would expect these data to actually be biased towards finding negative effects. Yet they don't find anything. One interpretation of this would therefore be that these are pure increases in homicides and no effect on actual property crime at all. That is, the increased homicides are not crowding out the reporting of property crime, and they aren't deterring crimes. If true, then it would suggest these are either making victims less safe (because they got killed and robbed anyway), or these are "mistaken identity homicides". That is, the victims are people who weren't going to commit a crime, but the offender believed they were and shot and killed them.
posted by scunning at 11:45 AM on June 14, 2012 [5 favorites]


Ardiril, I'm pretty sure that's a good definition of murder, rather than homicide. I believe that homicide -- strictly speaking -- includes the deliberate and lawful killing of one person by another, such as in war or capital punishment.
posted by gauche at 11:45 AM on June 14, 2012 [2 favorites]


Now that's what I call circular reasoning. Since homicide can be defined as "the deliberate and unlawful killing of one person by another"--without any regard for the act's legality, a law that strengthens self-defense will inevitably increase homicide

That's not circular reasoning at all. You've provided the definition for murder, not homicide, which is the killing of a human by another human. These are two different things. Stand your ground creates more situations in which a homicide might occur, justified or not.

It creates more killings.
posted by Ironmouth at 11:47 AM on June 14, 2012 [5 favorites]


This paper examines whether aiding self-defense in this way deters crime or, alternatively, escalates violence.

These aren't alternates.

On the other hand, we find that murder and non-negligent manslaughter are increased by 7 to 9 percent. This could represent either increased use of lethal force in self-defense situations, or the escalation of violence in otherwise non-lethal situations.

Wouldn't it have to be the latter? Increased use of lethal self-defense wouldn't be murder, right? Or are they saying that self-defense may have increased a ton and some people are overdoing it, or doing it wrong, and so spilling over into murder?
posted by DU at 11:49 AM on June 14, 2012


The critical thing is in the actual introduction to the paper:

Suggestive but inconclusive evidence indicates that castle doctrine laws increase the narrowly defined category of justifiable homicides by private citizens by 17 to 50 percent, which translates into as many as 50 additional justifiable homicides per year nationally due to castle doctrine. More significantly, we find the laws increase murder and manslaughter by a statistically significant 7 to 9 percent, which translates into an additional 500 to 700 homicides per year nationally across the states that adopted castle doctrine. Thus, by lowering the expected costs associated with using lethal force, castle doctrine laws induce more of it. This increase in homicides could be due either to the increased use of lethal force in self-defense situations, or to the escalation of violence in otherwise non-lethal conflicts. We suspect that self-defense situations are unlikely to explain all of the increase, as we also find that murder alone is increased by a statistically significant 6 to 11 percent. This is important because murder excludes non-negligent manslaughter classifications that one might think are used more frequently in self-defense cases.

So there's suggestive, but inconclusive evidence that justifiable homicide ('lethal self defense') increases, but aside from that, straight-up murder seems to increase as well (as far as the crime statistics are concerned).
posted by dismas at 11:52 AM on June 14, 2012 [3 favorites]


Where "Stand Your Ground" laws exist, the "bad guys" have a BIG incentive to act preemptively. In other words, shoot first. Good luck with that, Civilization.
posted by oneswellfoop at 11:59 AM on June 14, 2012 [9 favorites]


Well, if the laws lower the cost of using lethal self-defense, and do that socially as well as legally by making people feel that they've been authorized to defend themselves, they also raise the cost of committing crimes. If that rise doesn't translate into a lower property crime rate (and it doesn't seem to), then there should be a logical increased willingness for criminals to use lethal force themselves to hedge against the increased risk now involved in property crime.
posted by OmieWise at 12:00 PM on June 14, 2012 [2 favorites]


To put it less succinctly and clearly than oneswellfoop did.
posted by OmieWise at 12:00 PM on June 14, 2012


oneswellfoop put it more succinctly, but I also like your generalization via game theory. I wonder if crime stats have been studied that way rigorously.
posted by DU at 12:03 PM on June 14, 2012




Where "Stand Your Ground" laws exist, the "bad guys" have a BIG incentive to act preemptively. In other words, shoot first. Good luck with that, Civilization.


And to bring friends. And better firepower.
posted by Pogo_Fuzzybutt at 12:06 PM on June 14, 2012


DU: Wouldn't it have to be the latter? Increased use of lethal self-defense wouldn't be murder, right? Or are they saying that self-defense may have increased a ton and some people are overdoing it, or doing it wrong, and so spilling over into murder?

I think the authors are intentionally not drawing a conclusion, for whatever reason. I think it may be asking a lot of the data and the tests they can do to really say strongly that it's the latter and not the former, though I think the pattern of evidence taken at face value suggests it is the latter.

Imagine Florida with and without the law. Without the law, there are two people we'll call Offender and Victim. Assume that in this timeline, Offender robs Victim and Victim doesn't resist, giving us a baseline of 0 homicides and 1 robbery.

1. With the law, O tries to rob V, V stands his ground, O kills V and robs V. Robbery and homicide, but due to the hierarchy rule, UCR will only report the homicide since it is the worse offense. (+1 homicide, -1 robbery) because there's 1 more homicide but 1 fewer reported robbery (due to the hierarchy rule suppressing the robbery.

2. With the law, O tries to rob V, V resists and kills O. No robbery. (+1,-1). That is, 1 more homicide, 1 less real robbery, than the counterfactual.

3. With the law, O tries to rob V, V resists and stops O. No robbery, no homicide. (0, -1). This is the deterrence case. There was no homicide in the counterfactual, nor in this timeline. But there is deterrence, hence the negative one.

Now imagine a timeline where O doesn't try to attack V. Here the baseline is 0 homicide, 0 robbery.

4. With the law, V faces incentives to approach O, a fight breaks out, and V kills O. (+1,0). That is, one more homicide and no change in burglaries.

5. WIth the law, V faces incentives to approach O, a fight breaks out, and O kills V. (+1,0). That is, one more homicide, and no change in burglaries.

6. With the law, V faces incentives to approach O, and while there is a stand off, no one is harmed. (0,0). No change.

I could go on, but I think no matter the iterations, if you are seeing an increase in self-defense homicides, then it should be accompanied by decreases in robberies. Either because of the hierarchy rule which is actually a form of selection bias -- almost guaranteeing that you will see reductions even when they aren't there -- or you will see actual deterrence.

But they do not find this. They in fact are finding no effect on robberies, etc. The results are purely from the increases in homicides. Notice that the (+1,0) cases are coming from either the situations where the robberies would've happened anyway (numbers 4 and 5). If the homicides actually did break off an existing crime, then as far as I can see, it should reduce the reporting of the other crimes. But finding a zero effect on those others to me suggests that there's strong evidence that this is evidence for unnecessary homicides (for lack of a better word), and most likely cases of mistaken identity.

I suspect the authors are reluctant to try and make deductive conclusions like this, though.
posted by scunning at 12:11 PM on June 14, 2012 [5 favorites]


"Homicide" in the strictest medico-legal sense, is defined as: the death of a human being caused by the action (or inaction in certain circumstances) by another human being.

No value judgment is placed on the act, the actor, or the decedent when the classification is made.

Such judgments are deferred to the law enforcement/judicial arena.
posted by Renoroc at 12:23 PM on June 14, 2012 [1 favorite]


*waits in darkened porch of MeTa with loaded shotgun*
posted by ceribus peribus at 12:27 PM on June 14, 2012 [4 favorites]


Definition of murder and non negligent manslaughter in the UCR is: here:

Criminal homicide—a.) Murder and nonnegligent manslaughter: the willful (nonnegligent) killing of one human being by another. Deaths caused by negligence, attempts to kill, assaults to kill, suicides, and accidental deaths are excluded. The program classifies justifiable homicides separately and limits the definition to: (1) the killing of a felon by a law enforcement officer in the line of duty; or (2) the killing of a felon, during the commission of a felony, by a private citizen. b.) Manslaughter by negligence: the killing of another person through gross negligence. Deaths of persons due to their own negligence, accidental deaths not resulting from gross negligence, and traffic fatalities are not included in the category Manslaughter by Negligence.
posted by scunning at 12:35 PM on June 14, 2012


In related news, it turns out (at least in Texas) you can't stand on someone else's ground, shoot an unarmed person 20 feet away from you whose hands are in the air, record yourself doing it, and then claim a stand-your-ground defense.

The lesson here, of course, is don't record yourself committing murder and maybe you can get away with it.
posted by dirigibleman at 12:38 PM on June 14, 2012


The lesson here, of course, is don't record yourself committing murder and maybe you can get away with it.

Crazy thing is he thought the video would exonerate him. From his (deluded) perspective he thought he WAS just defending himself against agressive/criminal neighbors.

Humans are funny things. :-/
posted by LordSludge at 12:48 PM on June 14, 2012


Robert Heinlein would argue (and you cannot prove him wrong) that stand your ground is in dangerous places which would be more dangerous if they didn't have stand your ground. I disagree with this, but this study is not proof like E=mc^2.
posted by bukvich at 12:49 PM on June 14, 2012 [1 favorite]


bukvich -- Proof in theory comes from deductions based on basic axioms -- like E=MC^2. Evidence for a theory's predictions comes from a different process, though. Proof and evidence in other words are not the same thing.

Your point is a good one. If states who have these laws were "worse" (say their pre-treatment crime rates were worse) than non-states, you have to address that somehow. This is sometimes called selection on observables vs. selection on unobservables. The "differences-in-differences" links I posted was my effort to present some information as to how studies like these are specifically designed to answer questions like the one that you're raising.

So long as the thing that makes these states "different" and which made them more likely to adopt the law doesn't change over time, then this method actually eliminates that bias from the estimation by de-meaning the state outcomes and explanatory variables. In other words, if states with characteristic X were more likely to adopt these laws, then the first difference in this method deletes the X by comparing the same units over time. This requires the X be linear in the outcomes, though. The second difference then compares the differences of each unit to the other state differences.
posted by scunning at 1:07 PM on June 14, 2012


Crazy thing is he thought the video would exonerate him. From his (deluded) perspective he thought he WAS just defending himself against agressive/criminal neighbors.

The video. He clearly wanted to kill the guy. He wanted to start some shit and be able to finish it "I told you so" style. Bill Cosby talked about how he used to own a gun and do the same thing.
Cosby: "I had the gun in my pocket. The reason I had it was to protect my family. But I also knew that anything that went on outside - and it appeared to be something that wasn't on the okay - I went out with my gun. And the thought was - if this person was not right, or if that doesn't move when I say move - I'm going to show that I have a gun."
This guy's mistake was not realizing who he was shooting. If it was his scary black neighbor, he may have gotten away with it. Instead, he shoots a white elementary school teacher. Dummy. Who are these people who hang around when guns start getting flashed? What the hell.
posted by cashman at 1:11 PM on June 14, 2012


I've read the report, and isn't every single stand your ground case referenced in the report involve someone who was ruled to have had a legitimate fear of being killed or maimed? I believe then each of these examples of deadly force occurred in lieu of a rape, murder, kidnapping, aggravated assault or some other form of violence where the victim feared for their life.

If that's the case, then I don't understand why the increase in the use of deadly force is a bad thing. If this is justifiable homicide – a response to a lethal threat – then why is this a problem?
posted by lstanley at 1:48 PM on June 14, 2012


Since Florida adopted the first castle doctrine law in 2005

Interesting. I left Florida notably before 2005 but I was sure I'd heard it stated repeatedly that there was not a duty to retreat if you were in your own home. I suppose that could have been established by precedent not legislation.
posted by phearlez at 1:50 PM on June 14, 2012


I believe then each of these examples of deadly force occurred in lieu of a rape, murder, kidnapping, aggravated assault or some other form of violence where the victim feared for their life. If that's the case, then I don't understand why the increase in the use of deadly force is a bad thing. If this is justifiable homicide – a response to a lethal threat – then why is this a problem?

It is a problem because the whole point of this posting is that the findings of this study contradict your belief. The findings demonstrate that a rise in deadly force occurs not in lieu of but along side of other forms of violence. If the law operated as you suspect it does then you would expect to see a rise in homicides with a corresponding decrease in those other forms of violence you mentioned. The evidence does not bear that out where the "Stand Your Ground" law is in effect. That is the problem.
posted by jnnla at 2:16 PM on June 14, 2012


jnnla, I'm not following you because I didn't mention anything about a decline in crime rate. I simply stated that the increase in justifiable homicide is preferable those people becoming crime victims.

Yes the statistics prove that Stand Your Ground increases the number of justifiable homicides but does not prevent crime. That conclusion is accurate. This does not make the law a bad law in my mind, and in the mind of many others, because I don't consider justifiable homicide the worst outcome; they’re the second-worst possible outcome of a lethal-force situation. The worst outcome would be the the crime on the victim.
posted by lstanley at 2:40 PM on June 14, 2012 [1 favorite]


Let's also point something out. The place this is likely to be used is in cases involving two or more bad guys anyway. A large number of murders are drug-crime related. From what I've read, there are a lot of bad guys who are using this law as a defense to murder in cases where they are engaged in turf wars and the like.

A colossally stupid idea.
posted by Ironmouth at 3:00 PM on June 14, 2012


They in fact are finding no effect on robberies, etc. The results are purely from the increases in homicides. Notice that the (+1,0) cases are coming from either the situations where the robberies would've happened anyway (numbers 4 and 5). If the homicides actually did break off an existing crime, then as far as I can see, it should reduce the reporting of the other crimes. But finding a zero effect on those others to me suggests that there's strong evidence that this is evidence for unnecessary homicides (for lack of a better word), and most likely cases of mistaken identity.

We wouldn't expect to see a reduction in reporting if the shooting was preceded by a stalker confronting their target, or someone being chased by the driver of a car they had just cut off, right? I don't know what percentage of self-defense shootings this instances make up, it may well be insignificant, but I'm wondering if these possibilities are discussed.

-----

Read the whole thing

Where can the paper be read. The first link to NBER.org does not make it available.
posted by BigSky at 3:09 PM on June 14, 2012


Who are these people who hang around when guns start getting flashed? What the hell.

Clearly, it's the MAD magazine feature, SYG VS. SYG.
Seriously, I'd love to see this from somebody with the right art skills.
posted by dhartung at 3:37 PM on June 14, 2012 [1 favorite]


Interesting. I left Florida notably before 2005 but I was sure I'd heard it stated repeatedly that there was not a duty to retreat if you were in your own home.

Right. The duty to retreat is outside the home. Inside the home, you do not have that duty. The castle doctrine law changed the requirement outside such that you no longer were obligated to retreat from danger.
posted by scunning at 4:07 PM on June 14, 2012


Bigsky. Sorry my bad. I found the ungated version here at the authors homepage.
posted by scunning at 4:08 PM on June 14, 2012


We wouldn't expect to see a reduction in reporting if the shooting was preceded by a stalker confronting their target, or someone being chased by the driver of a car they had just cut off, right? I don't know what percentage of self-defense shootings this instances make up, it may well be insignificant, but I'm wondering if these possibilities are discussed.

Right. If the situation is only that someone is provoking the person, and not intending to commit a crime, then I agree. But then I guess the question is -- what is being implied about the stalker's behavior absent the law? For instance, say Florida hadn't passed the law. What would that same stalker have done? He would provoked the person, but not actually done any harm to the person? He was being provocative, in other words, but not actually doing anything criminal?
posted by scunning at 4:10 PM on June 14, 2012


I was watching football online, it was the ESPN US coverage, and had American ads at half time, and there was one for a gun shop, full to the brim with fucking guns. A proper ad, with guns in close up and everything. Fucking amazed me, honestly. I'm sure they said things like "It has never been easier to protect yourself" and talked about how cheap the guns were and (I may be misremembering but...) payment plans ffs! It just seemed sick to me, that you can advertise the means of instant death on tv. But then I'm some leftie European so what do I know.

No wonder there is so much fucking gun crime. Thank god we don't have them here.

(There was also an ad for an anti-depressant and it actually said "If you start to have feelings of suicide while taking Brand X..." - so gun shops and antidepressants which make you suicidal. Good luck America.)
posted by marienbad at 5:05 PM on June 14, 2012


Right. If the situation is only that someone is provoking the person, and not intending to commit a crime, then I agree. But then I guess the question is -- what is being implied about the stalker's behavior absent the law? For instance, say Florida hadn't passed the law. What would that same stalker have done? He would provoked the person, but not actually done any harm to the person? He was being provocative, in other words, but not actually doing anything criminal?

If the shooter can make a convincing case that they were in fear of their life at the time they acted, then it's self-defense.

I don't know whether or not, any and all behavior that can provoke a self-defense shooting are necessarily crimes. In some states, like California, making threats is a criminal offense. But if we ignore those states and situations like a violation of a restraining order, or anti-stalking laws, then I would guess (and I am not an attorney) that in some instances a stalker provoking a self-defense shooting might not have been doing anything that the laws would recognize as criminal, at the instant the trigger was pulled. The assumption though, is that if the shot was not fired the shooter would have soon been a victim of criminal violence.
posted by BigSky at 6:46 PM on June 14, 2012


No wonder there is so much fucking gun crime. Thank god we don't have them here.

I know this is classic gun-rights vs. gun control debate, but....

I don't know where you are from, but go ahead and compare your violent crime statistics to the US. I can pretty much guarantee you that if you don't have guns, you are going to find significantly more violent crimes and also more property crimes. Worse than some of the most crime ridden areas in the US.

I can't state that it is due to our gun laws, but chances are it is not disconnected.

Here are some graphs to back myself up.
posted by psycho-alchemy at 8:24 PM on June 14, 2012


A few things I noted after reading the paper (thanks btw for the ungated link!)

1. They state that English Common Law has a long-expressed "Duty To Retreat." This is one of the most misunderstood things practically ever. English Common Law had a duty to retreat to the wall. So essentially, you were required to try to back off to allow things to cool, and for you to back yourself against a wall in order to prepare to fight. This is also a time when violence was primarily done with swords. The advent of guns changed things, particularly in the feasibility of safely retreating. They reference this later, but only in terms of US law that came from English common law.

2. Five years is actually a really short time for a positive behavior shift large-scale. They cite that:
in Texas, "only 1.5 percent of adults age 18 and over have a concealed carry permit, and presumably only a fraction of those carry a gun on a regular basis.
So it is in fact staggeringly unlikely that the overall rates of burglary would shift in a statistically meaningful fashion without an increase in the amount of adults age 18 and over carrying concealed.

3. They are only measuring a specific few things here: burglary, robbery, and aggravated assault. They do not address assault or rape at all.

4. The increase in justifiable homicide is listed here as 17 to 50 percent increase, but only an estimated 50 a year nationally, which means the amount of these cases is increasing statistically significantly, but probably only because there was such a small overall set of cases found to be justified.

5. It cites an increase in murder and manslaughter by 9 percent, but fails to take into account the fact of judicial bias. While Stand Your Ground law may exist, it still has to be ruled on by judges, who may in fact not yet have internalized the shift intended to be realized by that law. If judges feel that cases should still be manslaughter, even if they fall under the grounds of stand your ground, they are still added to the statistical increase in murder and manslaughter.

6. It doesn't take into account the increased effects over time of having less individuals committing crimes or engaging in assault in the population.

7. They're using the UCR, which has notorious documenting reporting problems. In addition, the UCR counts justifiable homicide /only/ as those instances which killed "a felon, during the commision of a felony." So by the UCR, a criminal killed in the act of committing a lesser crime than felony would not be considered justifiable homicide.
posted by corb at 3:20 AM on June 15, 2012


Bigsky: If the shooter can make a convincing case that they were in fear of their life at the time they acted, then it's self-defense.

Of course. But that wasn't quite my question. I was asking about the counterfactual -- would absent these laws he have been in fear of his life? Would he have been provoked at all? The law has three parts:

1. You don't have to retreat in public
2. If you are acting in accordance of the law, there is no civil liability of a homicide
3. The prosecution in such cases faces a higher standard of evidence against you

These elements of the law create incentives for someone to go into an unknown situation in public where they *believe* there is a problem. Consider a situation where I see a young black kid out in public "up to no good, starting making trouble in my neighborhood" to quote Will Smith. I have a choice. I could retreat from the situation or I could approach him. I approach him rationally if and only if the benefits of approaching are greater than the cost. This law lowered the costs of approaching people. Why? Because one of the things that the law affected was the cost of "being wrong". Sometimes I see a black kid in my neighborhood and I draw the conclusion that he's up to trouble, when in fact he was minding his own business. And I remain around this kid instead of retreating because on the off chance I'm wrong, well I face little risk of penalty. The likelihood of some conviction against me for being wrong is lower now as a result of the law (see point 3) and in civil court I face no penalty (see point 2). Which means together there's always less of a case against me, lawyers are less likely to bring them, victims less likely to sue, and so on.

So let's say I do approach this kid. He's black, I'm white. What's going on through his mind when he sees me approaching him, who isn't doing anything? I think it's incredibly naive to not recognize the likelihood he will begin to also think it's him that needs to defend himself. To say otherwise is to ignore the entire American history of racial strife as well as countless studies in neuroscience on how badly we process external information when we are agitated, and even moreso how subtly race operates on our ability to make decisions and interpret situations in split second situations.

My point is therefore about the counterfactual. In the counterfactual -- that is, a world where Florida does not have this law -- some of these provocative situations never even happen in the first place where people have to defend themselves. Because in the counterfactual, if I approach someone and I end up shooting them, and I'm wrong, and the facts bear this out later, then I will face criminal and civil liabilities. That's what I think you're missing. You can move into a situation you never should've went into in the first place because of mistaken identity, create a problem that wouldn't have been there that escalates to the point where you are then required to justifiably defend yourself. This is where this type of study is focused. It isn't saying that these increased homicides are not cases of self-defense. It's saying for some reason, isn't it odd that you're finding these laws *cause* (assuming the conditions for identification of the causal effects are holding) an increase in murder and non negligent homicides, but do not decrease any property crimes or any other violent crimes.

Well doesn't that seem suspicious to you? If I saw the black kid in my neighborhood and I thought he was up to no good, and I hadn't approached him, then presumably were I right about him being up to no good, he would've stolen something, raped someone, assaulted someone -- something right? I mean presumably they are self-defending *because* this person was doing something criminal, right? Yet notice the findings -- the law has no evidence of deterrence anywhere. It's *purely* an increase in murders. That is not consistent with a story of self-defense where the person "standing his ground" had correct beliefs about the situation ex ante. In fact, it suggests the opposite -- that he had jumped to conclusions, formed erroneous beliefs about the situation, and escalated what was otherwise a benign situation and made it murderous.

Zimmerman may very well have been forced to defend himself against Martin, and the law may still have caused that boy's death by providing targeted incentives to Zimmerman to rush into judgment erroneously. That's where this breaks into a more complex problem -- how to assign moral culpability there is a challenge. But one thing I do know -- people have tremendous reasoning and cognitive biases. See Daniel Kahneman's work on this. One of them is the "jumping to conclusions' bias, and another is the "overconfidence bias". And not surprisingly, people perform strictly worse in both of them when the races of the so-called offender differs from the race of the person forming the belief. This happens all the time -- see Wolfers and Price Quarterly Journal of Economics article on how NBA referees were more likely to call fouls when the race of the referee differed from the race of player. This isn't always merely about deep-seated racism. It's also about the unreliability of memory and perception, and how that unreliability varies by the situation, our tendency to rely on quick thinking, and the relative races of the parties.
posted by scunning at 8:28 AM on June 15, 2012


corb: 1. They state that English Common Law has a long-expressed "Duty To Retreat." This is one of the most misunderstood things practically ever. English Common Law had a duty to retreat to the wall. So essentially, you were required to try to back off to allow things to cool, and for you to back yourself against a wall in order to prepare to fight. This is also a time when violence was primarily done with swords. The advent of guns changed things, particularly in the feasibility of safely retreating. They reference this later, but only in terms of US law that came from English common law.

But as their study is only focused on US crimes and the US law, is this criticism relevant? What was the US obligation in the non-Castle Doctrine states in their sample? The new laws did or did not materially change where the duty to retreat was outside the home or it did?

2. Five years is actually a really short time for a positive behavior shift large-scale. They cite that:
in Texas, "only 1.5 percent of adults age 18 and over have a concealed carry permit, and presumably only a fraction of those carry a gun on a regular basis.
So it is in fact staggeringly unlikely that the overall rates of burglary would shift in a statistically meaningful fashion without an increase in the amount of adults age 18 and over carrying concealed.
I'm not so sure you can say five years is a short time for behavioral shift to occur. There's a large literature on deterrence and people's responses to incentives in economics. Five years is a lifetime by these studies.

3. They are only measuring a specific few things here: burglary, robbery, and aggravated assault. They do not address assault or rape at all.

Interesting. I wonder why. Good observation. They have these data, and both of those are clearly situations they should be examining. My thought was the laws were focused on "car jacking" type crimes -- public situations of property crime. But if they are not focused on all violent crimes, then I agree with you.

I think aggravated assault is the only index crime reported in the UCR, though, of the assault categories. There is a separate file you can get listing *arrests* that gives you access to far more outcomes, but these arrests and not offenses. But rapes are available. Rapes suffer from under-reporting so it could be this is part of it, but I need to read the paper more carefully. I would still be interested in that outcome personally.

4. The increase in justifiable homicide is listed here as 17 to 50 percent increase, but only an estimated 50 a year nationally, which means the amount of these cases is increasing statistically significantly, but probably only because there was such a small overall set of cases found to be justified.

Good point.

5. It cites an increase in murder and manslaughter by 9 percent, but fails to take into account the fact of judicial bias. While Stand Your Ground law may exist, it still has to be ruled on by judges, who may in fact not yet have internalized the shift intended to be realized by that law. If judges feel that cases should still be manslaughter, even if they fall under the grounds of stand your ground, they are still added to the statistical increase in murder and manslaughter.

Correct me if I'm wrong, but these are recorded offenses. These aren't convictions. The Uniform Crime Reports and the Supplemental Homicide Report are counts provided by police agencies. They may not lead to conviction. So judicial bias shouldn't matter, should it? These are reported offenses only, and if I understand your point, judicial bias would only be relevant for convictions -- not the reporting of the event at all.

6. It doesn't take into account the increased effects over time of having less individuals committing crimes or engaging in assault in the population.

Why wouldn't it? They are finding a *net* increase in homicides. If the alleged attacker was going to commit 2 more homicides, and the alleged victim killed the attacker, then you would find -1 more homicides. So the difference in homicides is a net value -- it includes both the deterred homicides by successfully killing a would-be killer, as well as the homicide itself in its calculation. It's entirely coming from the ex post years of the law change in other words.

7. They're using the UCR, which has notorious documenting reporting problems. In addition, the UCR counts justifiable homicide /only/ as those instances which killed "a felon, during the commision of a felony." So by the UCR, a criminal killed in the act of committing a lesser crime than felony would not be considered justifiable homicide.

Can you provide a cite?

Also, the murder and non negligent homicide counts from the UCR suffer from systematic biases that are *correlated* with the passage of laws? That is, are you saying that states which pass these laws had *worse* reporting problems for murders *ex pose*? Because if it is simply national UCR reporting problems, then every state has it -- not just states that pass these laws. They are finding increases ex post in the diff-in-diff estimation for those states that pass the law -- (1) relative to their own murders ex ante and (2) relative to other states own changes in crimes for the same periods. This means you have to make the argument that states who pass the laws then see increases in murders because their reporting either got more accurate (i.e., they always had higher murders, but before the law they were suppressed due to measurement error) or got worse. But just random error in the murder variable itself and uncorrelated with the law's passage will not produce a coefficient on the law variable that like this. This requires a unique reporting bias that is correlated with the passage of these such laws for that assuming I understand your point.
posted by scunning at 8:40 AM on June 15, 2012


Consider a situation where I see a young black kid out in public "up to no good, starting making trouble in my neighborhood" to quote Will Smith.

But why is this the assumed scenario? You're taking a high publicity event and presenting it as though it is representative of the class. Zimmerman's shooting of Martin is the kind of story the media, not to mention race hustlers like Al Sharpton, love. The Florida law passed in 2005, a number of other states have similar laws. Wikipedia's article mentions that the case law in West Virginia has been congruent with "stand your ground" statutes since 1921. The lack of similar instances suggests that this scenario is more the outlier than the norm.

Even if there are more instances of trigger happy neighborhood busybodies than I'm assuming, I would still rather give the benefit of the doubt to the shooter who started a conversation with a stranger who then takes it physical, than to prevent him from using a firearm because he chose to have unfriendly, but non-threatening contact with this stranger. After all, if the neighborhood busybody, who has been encouraged to confront others by the "Stand your ground" laws, should threaten anyone they too have to deal with the possibility of the stranger legitimately defending themselves with lethal force.

These elements of the law create incentives for someone to go into an unknown situation in public where they *believe* there is a problem.

You exaggerate the change in incentives. First and foremost, most people don't want to kill anyone. There are also no guarantees the police and courts will come to the correct conclusion, assuming here that the facts support a finding of self defense. No one at all wants to create an opportunity for the justice system to make a mistake when they're the ones who have to pay. This is why most people go out of their way to not be subject to police attention. Even if they're found innocent, dealing with mistakenly brought charges is a huge ordeal - stress on children and other relatives, possible break of marriage, drain on savings, perhaps spending a considerable period of time in jail if they can't make bail. There's also a chance of retribution from the friends and families of the person killed. And if it isn't a clear cut case, stigma from the killing. I can't provide a citation, but some time ago I read that a substantial percentage of shooters in self-defense killings moved away from the area shortly afterwards. Are there people out there who are indifferent to these disincentives, and need criminal sanctions to restrain their aggression? Without a doubt, but let's not exaggerate their numbers either.

I understand your point about the counterfactual, but you are still looking at the worst case scenarios. There are also times and places where you very much do want people to go out, and to perhaps talk to strangers in their area. I don't think we want laws that encourage people to be yet more reluctant. Or at least in the case of someone dealing with a stalker, not to feel that the legal burden is so heavy that they can't retaliate without having shown their willingness to retreat first. Are you familiar with Ken McElroy's story (on MeFi)? And I'm not referring to his murder, but to his years of intimidating the locals. Some years ago I saw a TV show dealing with a similar case in California* where the shooter was arrested and convicted, but again the neighborhood had been terrorized by this bully for years, who like McElroy liked to lurk around his neighbor's homes and follow them when they went for a drive. "Stand your ground" laws do not provide an easy solution for these relatively extreme situations, and the laws will also have the occasional tragic consequence, but their absence makes it even more easy for bullies like this to do as they will.

-----

* - Is anyone familiar with this story? It would have taken place sometime between say 1985-2000, in a suburb. The bully looked to be at least thirty, unemployed and had started up immediately after moving in with his Mom. He was shot by a married retiree after he had followed him down a hill. I believe they were both in their cars at the time of their shooting. A number of the neighbors testified about the bully's actions, but the man had little claim to being under threat with each of them in separate cars, and he was sentenced to a number of years in prison.
posted by BigSky at 2:01 PM on June 16, 2012


I don't understand - of course the property crime rate wouldn't change right away. Instead the property crime drop would lag behind. Think about it - the only time the shooting is justified is if the Offender was already in the act of committing a crime. For example, if an Offender breaks into a Victim's house then the Offender has already committed Burglary. Before the Stand Your Ground law defending oneself with deadly force could've been construed as a criminal homicide which is something the Victim would try to avoid (at one's own risk). After the Stand Your Ground law the Victim would be safe in defending themselves after the initial crime has been committed (Burglary in this case). There would be a drop in property crime, however, after some time since these types of property crimes are usually committed by repeat offenders. Since the Offender, in these cases, would be killed there would be less property crime after a few years (barring any other factors).
posted by enamon at 7:26 AM on June 18, 2012


« Older Behind-the-scenes Makeup Photos from Prometheus   |   The Computer Tree Newer »


This thread has been archived and is closed to new comments