Firearms Policy and Evidence of Effectiveness

By Akiva Liberman :: December 21st, 2012


With the public discussion about gun laws taking place in the wake of the Sandy Hook Elementary School shootings, it’s useful to revisit a question I wrote about after the Aurora shooting in July: Are gun laws effective in reducing violence? That post explained why we don’t know—mainly, because we can’t do controlled experiments and because the crime data we have are too unreliable and incomplete.

First, one important observation. I think the policy discussion about guns and violence is the right one to have, more so than a policy debate about safety specifically in schools and how to make schools more impervious to such attacks. In Aurora, a mass shooting occurred in a movie theater; in Tucson, outside a supermarket where Gabrielle Giffords was meeting with constituents. We don’t want to treat each setting as a separate problem. The comparison to movie theaters also illuminates why a focus on locking down the schools is likely to have limited effectiveness: schools are small parts of communities that are largely open. Even if we rendered the schools impregnable, suicidal killers could reach many of the same targets outside of those locked-down schools. We cannot lock down the entire community. We should also remember that, as horrific as these events are, the vast majority of shootings of children happen outside of schools; less than 2 percent of youth homicides happen at school. The tremendous resources it would take to render all schools impregnable could be much more effectively used to reduce youth violence.

So, are gun laws effective? As explained in my previous post, we don’t know. I was part of a scientific panel that reviewed the evidence; we concluded that the needed scientific evidence we usually demand to gauge effectiveness is generally not available when it comes to gun laws. But not knowing is very different from finding that these policies are not effective.

Policy discussion cannot—and should not—await that evidence. There are other considerations, including suggestive evidence from other countries, to bring to bear. We will never find out whether a policy, such as limiting access to guns or making them less lethal, will reduce violence unless we try it.

Yet, even if we enact laws that could be effective, it would take four additional things for those effects to appear. First, passing laws is not enough; they also need to be funded and well-implemented. Even well-implemented laws restricting access and sale of some weapons would take a long time to show impact given the arsenal of weapons already easily available. Once those laws began affecting access to guns, it would take additional time to see their effects. Finally and fundamentally, it will also take a crime data infrastructure that does not yet exist.

So, we need to make policy now. Then we will certainly want to learn if the new policies are effective. But we must not be naïve in demanding quick evidence of effectiveness. Let’s couple any firearms policy changes with the funding and dedication to greatly improve our crime data, so that we can learn what works.


  1. Seven reasons not to mandate armed officers in schools  ::  4:29 pm on April 24th, 2013:

    [...] cannot be isolated from their larger communities—and we cannot lock down entire communities, as I discussed in an earlier posting. In many cases, the costs of deploying armed officers at schools could be put to better use for [...]