A new study of the effect of red flag laws on suicides was published this month. The study title and a first glance at the Methods section, got my hopes up.
The impact of gun violence restraining order laws in the U.S. and firearm suicide among older adults: a longitudinal state-level analysis, 2012–2016
“This is a longitudinal study of US states…”
I consider longitudinal studies of the effects of laws to be the gold standard. Look at a period of time stretching from well before the enactment of a law to well after. Look for discontinuities in the trend of whatever factor you are concerned with at the time of enactment and after. If there is no discontinuity, the law had no effect on that factor.
But then I considered the time frame.
“… using data from 2012 to 2016.”
That’s a peculiar time frame, given how few red flag laws were passed in those years. In fact, only four states had red flag laws at all in 2012-2016: Connecticut (passed in 1999), Indiana (2005), California (2014), and Washington (2016).
Only California could possibly show a potential trend discontinuity due to law enactment in the test period. What are they up to?
“We hypothesize that, relative to states with fewer firearm laws…”/i>
Aha! It isn’t a real longitudinal study; it’s the usual cross-sectional garbage. And we won’t even get into how they cherry-picked two age groups (55-64 and 65+) when the laws were imposed on everyone of all ages.
On to the researchers’ results.
“In unadjusted analysis, each additional firearm law was associated with a significant 0.13% decrease in firearm-related suicide among older adults in both age categories”
That’s significant? Florida saw a 8.5% increase in suicides after passage
of its red flag law (granted, the post-passage time frame is too short to know if that is significant, or just a statistical glitch for one year).
Clearly that wasn’t so significant after all…
“After adjustments, GVRO laws remained associated with a 2.5% decrease in firearm-related suicide among older adults”
Translation: the data didn’t say what they wanted, so they massaged it. And still couldn’t get results consistent with what happened in Florida. If they’d simply do a state-by-state temporal analysis of suicide trends, they wouldn’t have to adjust for demographics.
“However, there appeared to be a slight increase in non-firearm-related suicides in this age group”
How much of an increase? I don’t know, because the only data shown (Table 3) claims a decrease in both age groups: 55-64, -0.02; 65+, −0.35 in adjusted numbers. The unadjusted numbers aren’t given; maybe that’s where the increase was.
At best, given all possible benefit of the doubt, the researchers may have found a slight correlation between red flag laws and suicides. But, again, I don’t know; because the existence of a law doesn’t say anything about how often it’s used.
The paper notes that until 2006, Connecticut only used red flag orders to confiscate roughly 20 guns per year. Florida executed well over a thousand orders in its first year. And saw the increase in suicides that I mentioned.
I’d like to see state-by-state graphs of red flag orders served versus suicides. And even that might not tell us much as such orders may be related to threats to others, not self. Orders are commonly sealed from the public view; we know of an order, but not what type of threat triggered it.
But we likely won’t see such studies because I suspect they already know the data won’t support the victim disarmament narrative.