Let not efficacy stand in the way of feel good policies

Very interesting in an unintentional way.  From How a New York County Used the State’s ‘Red Flag’ Law to Seize 160 Guns by Andy Newman, Benjamin Weiser and Ashley Southall.  Two others helped research it.  Worth remembering - five journalists wrote this even before it was edited.  

This arises from some recent mass shootings and the perpetual desire that we should prevent these from happening.  A very human response.  

The challenge is that 1) we have a civil right to weapons and 2) no one has been able to show gun control laws have a material and reliably beneficial impact on gun deaths.  

Consequently, in recent years, there has been in some corners, a shift to so-called "red-flag" laws as the deus ex machina of gun control.  The motivation arises because in almost all mass shootings four things are usually true: 1) there is a history of mental illness, 2) there is a history of law-breaking or cruelty, 3) there is a dysfunctional family, and 4) the shooter is already known to law enforcement, a known wolf.

Red-flags are intended to break this dynamic.  The thinking is that if people can be diagnostically demonstrated to have a certifiable mental condition that predisposes them to harm themselves or others, then by red-flagging them, we can stop them from buying guns and therefore also stop them from committing mass murder. There are obviously several truck-sized gaps in the logic and reality of the argument but it is hard not to be sympathetic to the rationale.

However, it is not enough to want to appear to do good things.  The challenge is much, much harder.  We have to actually do things that are materially and demonstrably good.  That is hard work.  

I am inclined to want to believe that we could design a red-flag program which might indeed reduce gun deaths.  There are multiple challenges though.  Not least of which is that the evidence for gun death reduction is at best mixed.  For jurisdictions which have adopted a red-flag law approach, some report small increases in gun deaths, some small decreases and some show no change.

A social policy, especially one that butts up against a civil right, needs to have a clear and convincing evidence that the proposed law works.  We aren't there yet.

In addition, no social policy is uni-directional.  Yes, a red-flag law might theoretically reduce gun deaths.  It might also, by bringing the unbalanced into contact with police, increase police shootings.  Entanglement might cause a minor mental health issue to spiral.  A person with a mild mental health issue might also live in circumstances of high danger and low policing, thus putting their lives possibly at greater risk.  A red-flag order might imperil someone's ability to hold a job.  

In order to judge a new social policy, we need to consider the magnitude and reliability of its efficacy and balance that against possible unintended negative second order consequences.  And there are always unintended second order negative consequences.  

There are additional concerns beyond sheer efficacy.  History is replete with sustained and consistent evidence of mental health professionals supporting authoritarian regimes using "mental illness" as a means of detaining, imprisoning, and controlling their citizens.  It is a real and common pattern to be concerned  about.  

Mental illness is always a continuum and it is dramatically difficult to draw reliable and explicit hard lines.  Where does quirky leave off and unbalanced take over?

Mental illness origins and manifestations have great variability and one size rarely fits all.  An intervention that works for one individual might not for another.  Enshrining red-flags into law tends to circumvent due process and to violate Hippocratic oath.  The intervention, well intended, does not, can not, address the unique circumstances of the individual.  

Further, even if we do have great confidence in the diagnostic reliability of mental health workers, what are the due process protections to defend from false accusations or to protect confidential information so that it is not leaked and used as a scarlet letter weapon against individuals.

Red-flag is an ill-defined and amorphous social policy with no clear evidence to its efficacy and plenty of reasons to be mindful of its past immoral use by others.

So red-flagging is back in the conversation again with much passion, over-weening confidence and little evidence.  

What do the New York Times reporters have to say?  

If you do not read deeply, you can walk away from the article with the impression of honest social policy experimentation with the prospect of success.  But if you look for what is missing and what is blandly said, you have a different conclusion.

As always, context is missing.  This is New York and Suffolk County.  For each, what are the number of guns, gun murders, and gun deaths?  This gives us context to estimate efficacy.  We don't have this data in the article.  

A quick search and some back of the envelope estimations provide an order of magnitude context.  

6 million New York Guns -  28% of Northeast households own a gun, implying that (given 7.4 million households in New York) there are 2,072,000 households with guns.  Given that there are some 500,000 gun checks to purchase a gun in New York, this doesn't sound unreasonable.  Given that each gun owner owns on average three guns, this puts the number of guns in New York State at around 6 million.  Order of magnitude estimate.  

804 New York Gun Deaths  

349 New York Firearm Murders 

450,000 Suffolk County Guns - Suffolk is about 7.5% of New York States population so an order of magnitude estimate would be some 450,000 guns in Suffolk County.

NA Suffolk County Gun Deaths

 84 Suffolk County Firearm Murders

Now, with that as context, back to the article.  From the headline we know that 160 guns were seized in Suffolk via the red-flag law.  Out of 450,000 guns.  It does not seem that this law really does much given the number of guns out there.  

Lots of stories and narrative, but the article includes, scattered in among its 2.200 words:

Suffolk County deputy sheriffs typically serve the order and remove the weapons, walking into situations about which they know little.

“It can be very high risk,” said Chief Deputy Sheriff Christopher Brockmeyer. “We try to do our due diligence and vet the respondents as much as we can.”

[snip]

In Suffolk County, the heavier use of red-flag orders does not appear to have produced significant changes in gun death rates compared with those in the rest of the state. But Ms. Hart, the former police commissioner, said that the county saw several positive effects, including forcing parents to confront their children’s psychological problems.

[snip]

A state courts spokesman, Lucian Chalfen, said that by law, red-flag records must be sealed after they expire. But the Suffolk cases, many of which involved expired orders, were nonetheless found in a commercial legal database and, in some cases, open court records.

[snip]

Though Suffolk County judges grant most requests for orders, there are plenty of exceptions.

[snip]

“They would call me and I just wouldn’t go, because I just didn’t want him to lose his job,” she said.

Against all these negatives, there is a lone voice speaking to the red-flag policy as a demonstrably good thing.

But Ms. Hart, the former police commissioner, said that the county saw several positive effects, including forcing parents to confront their children’s psychological problems.

But that was not the intent of the law.  The law was not passed to have the state intervene and force parents to confront their children's psychological problems.  It was intended to reduce gun deaths and it does not appear to do so.

That a former police commissioner is focusing on citizens being coerced to do something that the police commissioner thinks to be a good thing, but which is not actually a legal obligation, is alarming.  Making mental health services easily available is a good thing.  Forcing people to use mental health services has a very dark history.

If we are to make decisions on public policy based on efficacy and balance of costs, the New York Times reports that the red-flag law as implemented in Suffolk County is 1) risky for the police to enforce, 2) that there is no apparent impact on the gun death rate, 3) that the data collected by the state is not secure, destroying citizen privacy, 4) judges seem to treat red-flag requests on a rubber stamp basis (recalling that this was the fate of the constraints supposedly put on the Federal Intelligence Surveillance Court which was supposed to protect Americans' privacy but which devolved into an automatic approval process), and 5) the red-flag process imperils the financial livelihood of those being investigated.

Despite the NYT reporters providing the evidence that this an ineffective and problematic public policy, all the top-rated reader comments are gung-ho for spreading red-flag laws.  They do not seem to have even assimilated what was actually reported.