Do waiting periods help or hurt?

The answer is “it’s messy,” but in short, they are ineffective.

Take-aways

• Only one of 16 study combinations showed any significant change.
• On one statistical measure, the effects in the other 15 combinations were likely random.

The Claim

Advocates claim that waiting periods reduce homicides and suicides.

The theory (which, as you will see, is wrong) is that some significant number of murders and suicides occur spontaneously. Advocates proffer that a person who gets angry or depressed goes to a gun store and then murders someone or kills themselves right away.

This assumption does not conform with what we know about most gun homicides and suicides.

In the United States, upwards of 85% of gun homicides are street gang related or nexus. Gun acquisitions by gang members are never via gun stores or (we suspect) very rare at gun shows. The Bureau of Justice Statistics notes that a rock-bottom minimum of 43% of all crime guns come from “street sources,” and the rate is obviously higher for street gangs. Most of these guns enter the underground markets from theft, primarily from cars parked at residences.

Suicides are rarely spontaneous, either. Most people depressed enough to end their own lives have had suicidal ideation for months or years before trying it. Among gun suicides, the rate among elderly rural males ¹ who have owned guns all their lives is astronomical.

All this aside, we chose to test the effects waiting period laws had on gun homicides, total homicides (regardless of means), gun suicides, and total suicides. It is important to check both gun and total rates for homicides and suicides due to “substitution of means,” a 50¢ phrase for “if I don’t have a gun, I’ll kill you/me some other way.”

But the data was messy…

The Surge of Sundry Statutes

The populace reacts in waves. Given the rapidly rising rate of violent crime in the 1970s and 1980s, the public reacted by throwing a lot of laws at the problem, at both the state and federal level.

This surge of statutes limits realistic research.

We needed at least four years of crime and suicide data before and after the enactment of waiting period laws to make a fair assessment. With CDC data only going back to 1979 ², we were constrained on older laws. Starting in 1993, 24 states with more than half of the country’s population passed habitual offender laws that had significant effects on homicides. This situation overlapped with the Brady Bill’s temporary five day waiting period that  applied nationally.

Combined, it leaves data from 1993 through 1998 ill-suited for analysis. After that era, only Washington D.C. passed a waiting period law with enough years of CDC data to make for a useful analysis… except that D.C. is such an oddball outlier on so many vectors that we declined to include it. D.C. has often been the nation’s murder capital (per homicide rates) and has the odd distinction of being surrounded by two other states with decidedly different sets of gun laws.

In other words, D.C. is a mess… often confirmed by C-SPAN televised sessions of Congress.

That left us with four states, each having four data sets to explore (gun homicides, total homicides, gun suicides, total suicides) totaling 16 things to score.

The Long and Short of it

In this study we are only looking at handguns, for two good reasons.

First, all state waiting period laws apply to handguns. This is untrue for long guns (rifles and shotguns). To avoid comparing apples to pineapples, we stuck to handgun homicides and suicides throughout.

The other reason is that handguns are the tool of choice for both homicides and suicides. Handguns are common for murder as concealment is often necessary, making it good for street gangs, the primary source of gun homicides in the USA. For suicides, it is all mechanical. Though notable people (e.g., Kurt Cobain) do exit the stage using legal-length shotguns, the dexterity or ingenuity to self-execute this way is physically awkward. Most people who want to die using a gun find handguns much simpler to employ.

Also note that in this study we only look at murder and non-negligent manslaughter, and not other elements such as negligent manslaughter and justifiable homicides.

What’s the Difference (in difference)

For each of the 16 measures we looked at both a “difference in differences”(DiD) measure as well as before/after slopes.

The “difference in differences” is a statistical method used to measure the effect of a specific change or intervention by comparing two groups over time. In this case we examine one state’s homicide/suicide data, where a waiting period law was enacted, against the average of all states that had no waiting periods during the same time. By focusing on the difference between the changes, it reveals whether the intervention had a unique effect on the group exposed to it.

The four states where waiting period laws were enacted, and where we had ±4 years of CDC data, were Indiana, Hawaii, Washington State and Tennessee. Southern, Midwestern, Western and a few islands in the middle of an ocean. That’s enough variety for anyone’s taste.

And The Loser Is…

Indiana was the only state that showed that the waiting period might have had an effect, but only on homicides; and even then, there are unusual factors at play.

Indiana’s gun homicide rate was already lower, and falling faster than states without waiting period laws.

Yet once those laws were enacted, at least for the first two years afterwards, their gun homicide rate went up more steeply and earlier than other states, which showed a more minor blip.

There is no concrete reason for this difference. One major and completely speculative explanation is Chicago crack.

This was the era of the arrival of crack cocaine, and Chicago was an import hub.

Nearby Gary, Indiana was a post-industrial hellscape. The G.I. Posse and Two Six Nation gangs operated an estimated 300 crack houses in the area. Hence, the initial wave of crack+gang violence that hit most urban centers may have hit northern Indiana hard and early. But this assumption is hard to quantify, and since Indiana is our only exception to the lack of effectiveness for waiting period laws, it is hardly worth digging deeper.

Though… well, we have to point this out: the slope of the line for Indiana’s gun homicide rates before and after are nearly the same. Before the waiting period law was passed, the slope was -0.3, and after, it was -0.2. Not much change, and the decline in gun homicides was not as good afterwards. Said differently, the downward trend after passage of the waiting period bill appears to be mainly a continuation of a prior trend.

Tennessee is a different story, but also with some oddities.

The DiD calculations indicate that the law led to marginal increases in gun homicide rates; but as they say, timing is everything.

Tennessee’s wait period law bumped a pre-existing three-day wait to 15 days starting in 1989. Recall that this was when violent crime nationwide was still sharply rising, later peaking in 1993. Even states without wait periods had rising violent crime. But the sharp 1990 jump in Tennessee’s gun homicide rate, following the activation of their longer wait period, by itself did nothing.

What is important is that on a statistical measure (p-values) in the “difference in differences” test, only Indiana (and only for gun homicides) showed the possibility of an effect. And as we saw with the two years immediately after the enactment of their waiting period, the cause/effect really isn’t there.

Worth the Wait?

All in all, there is hardly any evidence that handgun waiting period laws make a difference in homicides or suicides. One outlier (Indiana) is problematic due to other trends, and the other state with measurable changes (Tennessee) speaks against the efficacy of these laws.

As far as suicides are concerned, they have no effect on either gun suicide rates or total suicide rates.

If politicians in your state are contemplating enacting a waiting period to acquire guns, perhaps they should wait.

The p-value table

Since someone will ask, here are the p values for the DiD tests for all 16 evaluations.

When a p-value is below 0.05, then there is nearly no chance the effects are random. In other words, there is high confidence in the measurement.

Which makes all the others with p-values ranging from 0.19 (weak) to 0.78 (terrible) represent evidence of waiting period laws having no material effect. Indeed, for all suicide rates, the “strongest” p-value was a stunningly weak 0.47.