The NRA's proposal and Unpleasant Realities

Right now it is easy to pick on the NRA, but let's seriously examine their proposal - armed security at every single one of our nation's schools.  (I already took a brief look at this question).  There are over 130,000 schools in the United States, so a program to place a competent, vetted, armed guard would probably cost something along the lines of $10 billion (give or take a few billion) - but figure 100,000 security guards at $100,000 annually each (that would include administrative costs, training, vetting, health care etc).

A lot of money, but not an impossible amount if the problem is severe enough.  But is the policy in proportion to the problem?

There have been 137 fatal school shootings since 1980, taking 297 lives.  That comes to about 10 deaths a year.  But, for sake of argument, let us make several stipulations:

1. Armed guards at every school will absolutely prevent school shootings (more than a few observers have noted that there was an armed law enforcement officer at Columbine High School and he was unable to prevent or reduce the severity of the incident.)
2. School shootings are on the increase so let's assume about 30 deaths per year, not 10 (this year saw 37, and the worst year for school shootings - which included the Virginia Tech Massacre - saw 44).
3. We will also ignore some of the other costs of this kind of program - if you bring 100,000 new people into schools it is almost statistically certain that some will be child molesters or engage in some other sort of deeply inappropriate activity.

To prevent 30 deaths per year, the NRA plan would have us spend about $9 billion (for easy rounding).  That is  $300 million for each child saved.  By comparison it makes TSA seem like a bargain.  TSA only costs about $7.5 billion annually, but an airplane going down would easily cost 100 lives, so if they prevent just one air plane bombing a year they are saving lives at the comparatively cheap rate of $75 million per life (of course it hasn't been proven that TSA has prevented any airplane bombings...)

Children are precious, of course - but $300 million to save a single child's life seems like an enormous expense.  Obviously the numbers are estimates, but any reasonable changes would still result in an extraordinarily expensive program.  $300 million could easily fund some serious national efforts to promote and enforce firearm safety.  Since about 800 people per year are killed in firearm accidents, a program that made a modest dent in that number would be far more cost-effective.

(We could probably also save thousands of children's lives by investing the same amount of money in clean water initiatives in less developed countries - but that's another issue.)

Politics
There is a real danger to the NRA in making this argument.  After a massacre in Tasmania, Australia had a mandatory buy-back program for certain classes of weapons.  I argued that this would be too expensive to work in the United States.  But compared to the NRA's option it would be downright cheap - probably one year of the nationwide school police program would cover it.  So if they keep offering these kinds of policies they will probably find themselves isolated in policy-making.  (A cynical take is that the NRA's major funders are of course gun-makers who no doubt are drooling at the possibility of selling hundreds of thousands of guns to the new school security force.)

Policy
There are still no easy solutions to gun violence in the United States.  There are too many guns in circulation to prevent at least some bad people from getting hold of them.  But a laundry list that chips away at various aspects of the problem could make a real difference.  In practical terms such a policy might include:

1. Modest, realistic gun control measures (such as limiting access to assault rifles and high capacity magazines)
2. Enforcing existing gun laws (for example, studies show many gun dealers do not follow required legal steps in selling guns - perhaps an aggressive enforcement campaign could change that)
3. Studying and implementing best practices for school security
4. Serious studies of the links between mental health and violence
5. Studying the legal aspects of mental health and violence to better balance individual and community rights

I am by no means an expert on this issue - just a dope with a blog and some background in policy analysis.  Many of my readers (assuming I have any) will not like the politics of these proposals .  Others will be disappointed with their modesty - i.e. too many studies, not enough action.  But modesty is in order, particularly on the third leg of the stool, mental health.  It is not clear that the field is advanced enough to support large-scale initiatives for mental health screening.  Large-scale national initiatives will reap lots of false positives that will be expensive and ruin lives.  Before jumping on a National Mental Health program, some study is essential and part of that study should be on the legal questions of how to best handle the mentally ill who cannot make prudent decisions on their own behalf.

There is a tendency to over-correct in response to a crisis, but the over-reactions can have deep consequences of their own.  In the decade since 9/11 hopefully that lesson has been learned.