Some thoughts on gun control

Warning: this is a long one; and I'm very much in favour of gun control.

There's plenty wrong with Canada, but one thing we've got right (in principle at least) is a strong gun control.  Given the recent spate of shootings, both in Canada and the US, there has been a lot of discussion about gun control.  I'd tried to argue for strict gun control on Google+, but the arguments became scattered over several discussions and so may have lost some of their effectiveness.  So I'm going to try and put them all in one place here, in the hope that my position will make more sense.

To start, I'll summarize how I think guns should be controlled, just so that you know where I'm coming from.

While in a perfect world guns would be entirely unnecessary, we do not live in a perfect world; so the basic trade-off is the safety that guns can offer, in the right hands, versus the potential for disaster they offer when used otherwise.

Guns, like all weapons, were designed for only one purpose: to threaten and execute killing.  As such, they are abominable things, because to kill (for any reason at all) runs counter to our deepest instincts.  Unlike other weapons, guns live in a designerly "sweet-spot," being both maximally portable (and concealable) and destructive at short and medium range, as well as being cheap to make and buy these days.

I also distinguish between the cognitive and societal reasons for guns to exist versus their actual existence and use.  That is to say, I distinguish between the behaviour, the function, and the purpose of guns.  Behaviour is how a thing (or person) responds to stimulus; on this level, a gun just responds to an activation (pulling the trigger) by firing a very hard, accurately directed projectile at high velocity.  Function only arises from a thing's use in some context or situation, where the thing's behaviours are used by other things (or people) to achieve a goal.  Guns have many functions: to kill, to protect, to secure food, to frighten, etc.  Purpose arises from intent in a thing's creation, and distinguishes a given type of thing from other types of things; the purpose of a gun is unequivocally to kill, and only to kill.  Every other function of a gun is moot if not backed up by the threat of death.

So from a completely amoral point of view, guns are very well-balanced products (i.e. they are extremely effective and efficient).  Because of this, they remain the product of choice for achieving certain goals.

They are, however, not the only ways to achieve those goals, because goals are characteristics of agents interacting in complex systems.  And in all complex systems there are myriad ways of achieving goals.  Since there are many ways of achieving goals, one must consider guns, no matter how well-balanced they are as products, to be sufficient, but not necessary.  Furthermore, since the function of guns depends on the situations in which they exist, the decisions of whether to ban guns, what kinds of guns to ban or permit, and where (geographically) they should be banned and permitted, depend entirely on the situation.

I can summarize my argument as follows (please note that it is a general argument and not grounded in any constitution, American or otherwise):

1. Guns are intended to kill.  Even as deterrents, they're useless without the backing of a threat of death.


2. A gun-filled society is therefore one filled with the threat of death.


3. A society filled with the threat of death is not healthy.


4. We prefer healthy societies to unhealthy societies.


5. Therefore, one way to make a society healthier is to eliminate guns wherever possible.

And so we finally get to my proposal for gun control.  My proposal is not so much a set of rules as it is a method for making rules.  It doesn't provide answers; it provides a means for each society to determine its own answers.  Remember: rules depend on contexts, and contexts change.  So we want an effective and efficient way to set up rules and let them change to suit changes in context.

The method is quite simple:

1. Collect as much data on the nature of gun usage (including but not limited to acts of violence, hunting, target shooting, and military uses).  In particular, one must capture the class of weapon(s) used, the nature of the usage, the date of the usage, the "results" of the usage (crime, crime prevention, death, gun collection, securing food, protection from animals, protection from other humans, and so on) and the location of the usage.  One needn't go very far back in time; five years should be sufficient.


2. Create a series of maps showing the distribution of gun usage according to each of the other characteristics listed above.  The analysis will yield geographic areas where trends of gun usage emerge.  Some of these trends are already obvious, but sometimes even obvious things aren't necessarily correct.  The maps will validate (and in many cases correct) our informal perspectives.


3. Perform a cost/benefit analysis for each area with respect to each of the other characteristics.  The goal here is to identify whether each type of gun is of net benefit or net harm in a particular area.  This step could be very difficult to execute because this is where individual perceptions of the "value" of gun usage will appear.  But at least, in this method, this is the only step that is socially difficult.  For instance, say rifles are allowed in rural areas known for attacks by wild animals.  Rifles will help protect humans from those attacks, but they will also increase the chance of people dying (by being accidentally or intentionally shot).  If more people are saved from wild animal attacks than are killed via rifle usage, then it rather makes sense to allow rifles in those areas.


4. Outlaw gun types that pose net harm in each geographic area and according to the analyses described above.

The method should be repeated at regular intervals - perhaps once every five to 10 years - to ensure that all the latest data is driving the identification of gun types and areas where restrictions and bans are necessary.  It may be that the boundaries between regions change with time as populations move around.  Or it may be that new types of guns are introduced but become common only in some areas.  Whatever the case, the rules will be useless and best and harmful at worst if they are not defined correctly.  And since the context in which the rules work will certainly change, the rules themselves must be tweaked regularly.

Note that some factors are intentionally left out of my method.  For instance, whether guns are stored securely is irrelevant.  In regions where usage of a type of gun is correlated to net harm, then that type of gun must be banned, regardless of other factors, because net harm (or benefit) is the only meaningful criterion.

I lack the information (and the time and other resources) to execute this method.  But just to suggest the kind of rules that might emerge from its execution, here's one possible set.  Three broad types of geographic areas are identified: urban, suburban, and rural/remote.

1. The military and police can carry and use weapons in any geographic area, subject to existent laws and regulations.  Changing those laws and regulations is of course possible, but doing so falls outside the scope of my proposal.


2. You don't need an Uzi to kill a bear that's raiding your food stores, and you don't need a AR-15 to stop a rabid coyote.  A wide assortment of high-power, high-capacity guns must be restricted everywhere to only the military and possibly SWAT police teams.


3. All guns, regardless of type, must be registered.  All gun owners must be licensed, and a graduated licensing system for gun owners must be put in place based on the power and complexity of the gun.  All unregistered guns must be destroyed immediately. Any gun found in the possession of a person without an appropriate license must be immediately confiscated and destroyed.


4. In urban areas, no one can own or use guns of any type (exception: Rule #1 above).


5. In suburban areas, where the population and the police may be more sparsely distributed, and where encroachment by dangerous wild animals from nearby rural areas can happen, it may be necessary to allow some people to have certain types of weapons, subject to the rules below.  Certain allowances for the cultural practises of aboriginal people in North America may be necessary, but only in cases where practises involving guns are demonstrably connected to cultural ritual that does not conflict with other laws, and remain subject to Rules 1-3.


6. In rural areas, where both (a) citizens and police are even more sparsely distributed, and (b) dangerous wildlife is more easily encountered, a number of guns should be permitted by anyone who has had proper training and is properly licensed, subject to Rules 1-3.

Note that I say nothing about the criminality of people who are found to be unlicensed gun owners, or owners of unregistered guns.  The issue of what people do with guns is an entirely separate matter from having or not having the guns.  My only concern here is about the availability of the gun itself.

I also avoid dealing with hunting.  That's because I personally find hunting so barbaric that I cannot consider the matter dispassionately.  Some other group would have to consider how to deal with hunting.  So long as their decisions are based on evidence, then I'd be satisfied.

That's it.  That's my proposal.  Yes, it needs to be fleshed out (e.g. what constitutes "proper training"?) but those are details that can distract from the overall structure of the proposal and would require more in-depth study.

One may raise a number of concerns and counter-arguments about my proposal.  Some of them are discussed below.  (Most of these arose from various posts and subsequent discussions on Google+.)

Guns don't kill people, people kill people.

Actually, this is an oversimplification.  It's the person-gun system that kills people.  Guns make killing easy and cheap, and significantly detach the shooter emotionally (in many cases) from the victim.  Consider the difference of shooting someone and watching them just fall down dead, as if a switch had been turned off, versus pushing a knife into someone's chest, feeling the blade push through skin and flesh and muscle, rub against bone….  A gun sitting on a table somewhere will not kill anyone by itself.  But many people who might kill someone with a gun simply won't if they don't have it.  And then there's the question of accidents.  A gun sitting on a table won't accidentally go off and kill someone.  It takes a person and a gun to make that happen.  Remove the gun and the accident doesn't happen.  Removing guns lowers death and injury rates.

If guns are banned, then why not cars and planes too?  This argument hinges on cars and planes (and other devices) leading to injury and death.

This argument is flawed for two reasons.  Firstly, cars and planes are not intended as weapons.  That is, their purpose is to make life better, not extinguish it.  We already have various mechanisms in place to separate persons from cars and planes if those persons are likely to use the cars for unintended purposes that lead to injury and death.  We license drivers and pilots and register cars and planes.  Why not guns too?

Secondly, just counting injuries and deaths attributable to cars and planes - and comparing them to the corresponding rates for guns - is facile and naive.  To properly compare guns with cars and planes with respect to injuries and deaths, you need to look at the rate of injuries/deaths with respect to the rate of beneficial usage.  That is, for every injury or death caused by a person in a car, there are likely of instances of persons not being injured or killed by some other car.  That is, for all instances of car usage, only a tiny, tiny fraction of them lead to injury and death.  Now perform the same calculation for guns, and you'll see a very substantive difference.

"It's already illegal to kill people.  The murderers aren't really concerned with which laws they may be violating."

Sorry, but so what?  A person may intend to kill but won't actually do it unless the person finds the means and opportunity.  The ready availability of guns lowers the threshold for both.  Guns are, as I mentioned above, cheap, easy to use, highly effective, and highly efficient.  They literally constitute their own category of means of killing.  Furthermore, because they can be easily concealed and can be used very quickly, they open whole new regimes of opportunity.  By removing both the means and opportunities to kill that guns afford, injury and death rates will go down.

Training can fix anything.

Some people don't really see a problem, because they're responsible individuals who keep their guns safely stored all the time, have had extensive training, and are not the type to act stupidly.  They figure if they can do it "properly," then so can anyone else, and indeed that everyone has that "right."

I find that incredibly superficial.  First of all, while it's not necessarily objectionable for someone living in a rural/remote area to have a few rifles or shotguns for personal defence and hunting, this doesn't translate to letting city dwellers carry Uzis or concealed automatic weapons loaded with armour-piercing rounds.

Second, training isn't enough.  One must also have the strength of character to understand the significance and responsibility that comes with owning firearms.  We don't measure strength of character.

Thirdly, what about accidents and suicides?  (I provide some data below.) Training isn't going to stop someone from killing themselves - indeed, training might make it easier.  Suicide is usually indicative of underlying mental distress.  Do we really want to facilitate someone killing themselves when they could quite probably get some help and live happy and meaningful lives?  (The correct answer, in case anyone is wondering, is "No.")

Let me quote from Wikipedia on US gun violence: "There were 52,447 deliberate and 23,237 accidental non-fatal gunshot injuries in the United States during 2000.  The majority of gun-related deaths in the United States are suicides, with 17,352 (55.6%) of the total 31,224 firearm-related deaths in 2007 due to suicide, while 12,632 (40.5%) were homicide deaths."

I will not sacrifice or endanger that many other people for the sake of my own safety.  Because if I do, then I must expect everyone to also, and that will end up doing me even more harm.

Fourthly, this kind of evidence is anecdotal.  How many accidental or misadventurous deaths and injuries are those people willing to accept as a reasonable trade-off for the safety that they themselves feel?

Rights trump everything else.

This seems a particular issue in the United States, and it's odd that Americans continue to extol the virtues of the Second Amendment of the US Constitution (the right to bear arms) in light of how many other countries seem to be doing just fine without any corresponding legal scaffolding.  Perhaps Americans need to reconsider if they have finally evolved beyond the need for such constitutional rights.

A right to what anyways?  The right to own and use a device intended for no other reason than to kill people at short distances.  Wow.  What a fantastic right that is.  Wouldn't such a right be needed only if there are known risks that might reasonably be addressed only with guns?  Why would you need such a right if no such risks exist?

Most reasonable people will of course say that it's not about the right to own a killing machine; it's really about a right to protect oneself.  That is, it's not about the behaviour of the gun itself (killing), it's a question about the function that that behaviour embodies in a certain context (protection).

The problem persists, though, even if we recast it as a right of self-protection.  Such a right engrained so deeply in a constitution rather begs the question of which risks exist that providing protection is so important that it must be utterly embedded into a society.  If these risks really do exist, and if guns are the only legitimate way to provide protection, then countries with strong gun control should be rife with crime.

But they're not.

Here's some interesting, and typical, information about gun related crime in Canada and the US at about.com.  (Similar statistics exist for many other countries.)

• 1 in 4 Canadians have a firearm, and 1 in 25 Canadians have a restricted firearm (mostly handgun).  In the US, 2 in 3 Americans have a firearm, and more than 1 in 4 Americans has a handgun.


• For 1987-96, on average, 65% of homicides in the U.S. involved firearms, compared to 32% for Canada.


• For 1987-96, the average firearm homicide rate was 5.7 per 100,000 in the U.S., compared to 0.7 per 100,000 for Canada.


• For 1989-95, the average handgun homicide rate was 4.8 per 100,000 in the U.S., compared to 0.3 per 100,000 for Canada. Handguns were involved in more than half (52%) of the homicides in the U.S., compared to 14% in Canada.


• For 1989-95, the average non-firearm homicide rate was 3.1 per 100,000 people in the U.S., compared to 1.6 per 100,000 for Canada.

The article also goes on to show that on average, Canada is much less densely populated that the US.  That's only partly true, because the averages are calculated using total land area.  In Canada, however, most of the population lives within 200 km of the US border.  If one re-calculates the density only for the densely populated part of Canada, one ends up with an average population density very close to the American average.  In other words, population density does not explain the differences between the US and Canada.

If you're worried about comparing the USA to Canuckistan*, then consider this: "More Americans are killed by guns each year than all other developed countries combined, spree killings appear to be on the rise, and unfortunately the Aurora shooting appears to be another among too many cases where mentally unstable people too easily acquire extraordinarily deadly weapons." (source)

There are also some very pragmatic concerns associated with providing a right such as that of the American Second Amendment.  To enable this right to everyone - which is what rights are all about, of course - we are required to create whole industries, lobby groups, supply chains, etc.  These are massive and complex systems.  And like all systems, they're full of leaks - inefficiencies that allow materials flowing through the systems (guns, gun parts, bullets, etc.) to literally fall out.  These leaked things will constitute a significant portion of the illegal guns in circulation.  In other words, implementing this right itself is a self-fulfilling prophesy by creating the risk from which it seeks to protect one.  Armed criminals is an unavoidable result of giving the right to bear arms to good and responsible people.

Another pragmatic concern is that the context in which the American Second Amendment was created.  Back then, guns were for the most part unwieldy, single-shot affairs with limited range, accuracy, and power.  Reloading could take a lot of time, and they were nearly impossible to conceal.  Also, the times were different; there were significant risks to people.  Modern life, and modern weapons, are very, very different.  Pretending that somehow the Second Amendment is still relevant in its original form today is the worst kind of dogmatism.

And that's assuming that what we're really talking about here is in fact a right to protect oneself.  Perhaps it's not protection we're after but rather safety; the right to be safe.  That's both much more consistent with rights as identified in other cultures, but it's also more fundamental.  One protects oneself for the sake of being safe, not the other way round.  Could it be that the Second Amendment was just a right to safety intended to address specific risks endemic in the society of the time?  Could it be that, if the American Founding Fathers were alive today, they might have phrased it differently?

(I will leave to other posts the question of whether we have rights at all.  Spoiler: I personally tend to agree with George Carlin, who proposed that there are no rights but only privileges that we bestow on each other.)

Not thinking in systems.

This is a generalization of a number of comments that I've seen.  Here's some examples:

• "We will only ever be treating the symptom of an underlying condition."


• "I think we have many many more problems in this country that we should focus on, such as education and poverty of our young."


• "Of course reducing the number of guns will reduce gun violence.  But it does not reduce violence in general.  That is the root of the problem."

All these ideas are founded on the incorrect notion that there is a hierarchy of cause and effect.  If such a hierarchy existed, then finding and addressing the "root cause" at the bottom of the hierarchy would cascade up the hierarchy causally, and a whole bunch of problems would essentially fix themselves.

A far more accurate and realistic way to model causality is via systems thinking.  A society can be modelled as a collection of agents, each interacting with various other agents in the system in non-linear, time-dependent ways, such that any agent's actions depend (in part at least) on the actions carried out by other agents.  In such a model, causal loops form.  Agent A does something that impacts Agent B, causing B to do something that impacts Agent C, causing C to do something that impacts A again.  Loops are generally of one of two types.  Reinforcing loops cause interactions between agents to strengthen (think of bringing a microphone too close to a speaker); dampening loops cause interactions between agents to weaken (think of how your toilet tank refills itself slower and slower as it becomes full).  In systems models, gun violence becomes an emergent property of the system as a whole and not of any one of its elements.

Here's a sample reinforcing loop for gun violence.

1. Guns can be thought of as a sign of distrust - in other people, in the "authorities" who are supposed to keep the peace, in the government's ability to maintain a lawful citizenry, etc.


2. Politicians, seeking power above all else, will play into that distrust, magnify it, and enable it by setting policies, making laws, etc. that assume the distrust is founded, regardless of the evidence.


3. Lobby groups, like the intellectually destitute NRA, use scare-mongering and lies to seed further distrust.


4. This reinforces the distrust in the citizenry.  Increased distrust will lead to more shootings, both accidental and intentional, because everyone will be fearful of everyone else.


5. The psychological stress resulting from all this distrust makes people angrier; they start to withdraw into groups of sufficient similarity to feel comfortable (a perfectly natural response) and start seeing "the other" as ever more a threat to their safety.


6. Certain susceptible people will succumb to paranoia based on ignorance, resulting in sentiments like "We live in a dangerous world.  Don't look to others to protect you."


7. Distrust increases in the citizenry as a result, making people feel unsafe.


8. To regain a sense of safety, citizens buy more guns, call on lawmakers for greater liberties in their use, and tend to vote for politicians that echo their fears, distrust, and anger.


9. More guns lead to more gun violence.


10. Go To Step 1.

I don't know if this loop actually exists, but it seems reasonable.  There are many - probably hundreds - of other loops too, intersecting with this one, dampening or strengthening elements in it.  For instance, gun control would form a loop that would intersect with the one I sketched above at issues of distrust.  Depending on how well the gun control is implemented, it can either lessen or increase distrust.  If there is visible evidence of a decrease in crime as a result of gun control, then this would, over time, lessen the distrust that citizens feel.  If it's done poorly (as in the infamous Canadian Long Gun Registry), then it can actually increase distrust - poor gun control means the politicians charged with maintaining public safety are incompetent, which causes citizens wonder if they need to protect themselves rather than expecting the government to do it.  But protect themselves from what?  Without information, the citizens will increase their distrust of everyone else in their community.  As they say, better safe than sorry.

A sidebar on point #6 above: "We live in a dangerous world.  Don't look to others to protect you."  That's an actual quote from someone who quite rabidly attacked my claims that gun control will lower the rate of violence.  On this particular counter-argument, I must call bullshit.  Cavemen lived in a dangerous world; we (that is, the "we" in the developed world who can afford to go to movies and buy assault weapons) do not.  We live longer and better, on average, than at any other time in human history.  Ours is not a dangerous world - except for the paranoiacs who think it is.

And with regard to "Don't look to others to protect you...." I ask why not?   After all, that's what society is all about.  That's what it's for - because together we can live better and safer than we ever can alone.  We evolved to be social animals.  That means that socialization helps ensure survival.

But just to underscore the depths of - I can only think of the word sociopathy to describe the sentiment - here is the entire quote, complete with errors: "Don't look to other to protect you... Why not.  Because the can't.  If you think the police are going to be there to help you you're mistaken.  They will clean up the mess and prosecute the guilty, if your lucky.  Don't want to die, then protect yourself.  Yes, shoot back.  If someone had shot him after his intent had been clear many would be alive.  Don't shoot 'at' him, don't just 'shoot back'  learn how to protect yourself so that when you do shoot you are not throwing lead in a general direction and hoping for the best but you actually know what to do.  Learning how to protect yourself is more then knowing where to buy bullets and how to pull a trigger.  It actually takes effort.  The payoff when some asshole in a theater or mall starts killing people you don't say boy I hope someone comes and protects me. No you defend yourself and the ones who depend on you for safety."

People who honestly believe this get to vote.  They are the ones who will elect the politicians who will make matters worse, as I described in the feedback loop above.

In my view, guns are both symptom and cause, along with lots of other factors.  To alter this system, one needs to find a spot somewhere in the system where a gentle tweak or poke in the right direction will cause those loops to alter the system for the better.

Of all the places one can poke the system, I've never found a better spot than severely limiting the availability of guns.  Doing so should radically alter the so-called "gun culture" and move society away from the reinforcing feedback loops that are currently operating in it.  And if a better spot can be found as a result of further research and systems analysis, then so much the better.

As things stand today, we understand relatively little about complex systems, and the citizenry is entirely ignorant of the field.  An understanding of these systems can be achieved, but only with massive efforts.

It turns out that the human mind cannot really deal with more than one feedback loop at a time.  You can read one of Dr. Terence Love's paper on this to find out more.  In particular, that paper covers the issue of developing interventions to lower crime rates - not a particularly disconnected issue from that of gun control.  This shortcoming of the human mind can be offset by developing models of complex systems that explicitly identify all the feedback loops.  That way, we can study each feedback loop, and how they interact, in a more structured way.

Just to emphasize how entirely inadequate the human mind is, look at this fantastic diagram of the principal influences affecting obesity, and keep in mind that we can only really internally treat one feedback loop at a time.  That influence diagram lets counsellors and therapists study, for any given case, which feedback loops matter in that case, and how best to intervene in each case.

Here's an example of the potential influence that a non-obvious factor can have: that just having a particular object can influence one's behaviour:

There's an argument that it doesn't matter whether assault weapons are legal or not, because in the case of the Colorado cinema shooting, the shooter's assault weapon jammed.  This is, of course, an instance of the "Cherry Picking" fallacy; that is, single instances where an argument seems to fail is not by any means proof that the argument fails universally.

But even if we set that aside, there's a system dynamics issue.  Perhaps just owning that assault weapon was enough to influence the shooter to go through with his horrific plan.  Perhaps it's irrelevant which weapon the shooter actually used at the cinema; just having the assault weapon may have changed his mental state enough to drive him forward.  He may have believed that the assault weapon would give him such an advantage that his attack would be "successful."  It may well be that if he had not had that particular weapon - whether it jammed in use or not - he may not have gone through with his terrible plan.

Yes, this is hypothetical.  But I think it's a good example of the indirect influences that firearm possession can have.

The power of evidence.

Here's an interesting image from an article by Richard Florida, published in 2011.  It describes correlations between gun deaths and various other factors for which data was available at the state level in the US.



This really helps show how interrelated so many of these factors really are, and how a systems approach is really the only way to grapple with the problem of gun violence.  David Friedman produced a very informative essay covering a number of aspects of the connection, if there is one, between drugs and (gun) violence.

The bottom line here is that I've not found any compelling evidence connecting gun violence with drug dealing or usage, over whole societies.  (There may well be localized correlations, however, in specific areas - e.g. specific cities.)

Some have argued that blanket bans are ineffective and point to the bans on addictive drugs as an example.  There's three problems here.  First, it's a straw man argument: no sensible approach to gun control includes a blanket ban.  Second, guns aren't addictive, so comparing guns to drugs is comparing apples and oranges.  Third, blanket bans are never (in my experience) part of a systems-oriented solution; therefore, they are highly unlikely to work.

Finally…

That's all I've got to say on the matter.  I'm not entirely sure I'm right about this, but none of the arguments I've heard so far have dissuaded me.  So I will continue to advocate for evidence-based, strong gun control, because it will lead us all to a better, healthier, and safer society.

---

* Dear Pat Buchanan, go fuck yourself.