“Universal” background checks are anything but universal. Criminals don’t bother.

Which is why we see only modest changes in some states that passed such laws, though the average of all such states shows a tiny increase in gun murders.

That tiny increase was significantly lower than in states without such laws.

Take-aways

• A small net increase in gun homicide rates (all things being equal) after passing universal background checks
• A tiny but larger increase in the same period for state that lacked such laws
• There were significant, concurrent anti-crime initiatives during some enactment periods which muddies the picture

The Concept and The Reality

“Universal” background checks impose a requirement on private gun buyers and sellers to conduct a background check on the buyer as would occur in a gun store.

This is a case of “nibbling around the edges.” The Bureau of Justice Statistics tells us every decade that most crime guns (43% in the latest tally) come from “street sources.” These are criminal exchanges typically between prohibited persons (i.e., convicted felons) and completely off the radar. These folks make a living disobeying laws, so the likelihood of them rolling down to the nearest gun store to run a background check is remote.

Notice that the same report lacks a line for “bought from a stranger” or “private transaction.” That means such otherwise legal transfers are a subset of the 5.9% of crime gun sources labeled “other.” This ignores transactions between friends and family, which are 8% of crime gun sources. But mixed therein are “wink and nod” known felonious transfers, and there is no breakout of those subgroups who would choose to ignore any “universal” background check law.

Summarized, “universal” background checks only apply to a fraction of a fraction of all gun acquisitions. As such, you cannot expect them to do much.

But we didn’t expect to see gun homicides go up.

The Statistical Summary

There are two types of “universal” background check laws: those requiring background checks for just handguns and those requiring such for any type of firearm.

We looked at every state that had enacted such laws that had at least five years of gun homicide data before and after enactment. The short list included Connecticut, New York, Delaware, Oregon, and Washington State. Immediately we see a tight cluster of Northeastern states that share similar cultural/political notions, and two Western states that don’t (we excluded California and Rhode Island, as their “universal” law was in place before the peak year of 1993 in violent crime and thus we could not obtain a faithful before/after picture).

For each of these states, we then compared their gun homicide rates with the average of all states having no “universal” background checks.

We’ll look into some of these states in more detail later, but the average for all the states’ gun homicide rates held two surprises: states that passed “universal” background check laws went from a flat gun homicide situation to a slight increase and states without any “universal” background checks were experiencing decreases beforehand and steeper climbs thereafter.

One plausible explanation was that four of the five states enacted their laws roughly at the same time (2013 through 2015). Nationally, gun homicides started rising strongly in 2015. Hence, states lacking “universal” background checks took the full brunt of crime gun acquisitions from any source.

One unsatisfying aspect of this explanation is that we know new restrictions drive crime gun acquisitions underground. A different BJS analysis conducted before and after implementation of the National Instant [Criminal background] Check System (NICS) showed that 3% of crime guns migrated from gun stores to street sources. One would think “universal” background checks would do the same unless the law was universally ignored. Hence, the explanation has weaknesses.

Using a difference-in-difference test we see that perhaps 14% of the lower gun homicide rates in “universal” states might be associated with the law, all things being equal. But they never are, as shown by the fact that the state-level difference-in-difference scores were wildly different, and in one case the entire degree of change could not be accounted for by the law itself. For example, Connecticut’s difference-in-difference score was 10% and Washington State’s was 53%.

The Connecticut Conundrum

Connecticut passed their “universal” law in 1999, a mere five years after the peak violent crime rate throughout the United States. Like 23 other states, they passed a habitual offender law (a.k.a., three strikes) in 1994, which was the starting year of our analysis for them. They also plowed $9.7 billion into prison expansion and grabbed $6.4 million in federal Community Oriented Policing Services (COPS) grants in 1995 alone. In short, Connecticut was on an anti-crime binge throughout the study period.

Before enactment, their gun homicide rate was falling on a slope of -0.4, but after their “universal” background check law was enacted, the rate of decline in gun homicides dropped only at a -0.1 rate. It is likely that in the “before” period, the sundry anti-crime initiatives were getting thugs off the street and causing the gun homicide rate to fall accordingly.

It is tempting to remove Connecticut from the amalgamated calculations, but we’ll leave it in for completeness’s sake.

The Delaware Ding

The first year of Delaware’s study period saw a sharp and temporary spike in gun homicides. Such small-population states often have sharp swings in crime data, as it does not take a large number of new crimes to create a large percentage shift. Delaware is #45 in size of state population; and in the year they passed their “universal” background check law, they had fewer than a million residents, about the size of Austin, Texas before it started growing faster than Willie’s Weed.

In our analysis we excluded that first year. But that did not help erase Delaware’s special status as the only “universal” state that had [a] a rising gun murder rate before the law and [b] a higher rate after (slopes of 0.1 and 0.2, before and after, respectively).

Leave It To Others

The non-Connecticut states were nicely clustered in their year of passage of “universal” background checks, so they should have some homogeneity, but they don’t. That said, the remaining states pit Delaware and New York (Northeast) against Oregon and Washington (Northwest). These East–West extremes are about as different as Vin Diesel and Dylan Mulvaney.
What we see in terms of change  in these four states is educational. Regardless of region, the gun homicide slope went from neutral before “universal” background checks were enacted to rising, but only slightly. Yet in contemporary timeframe states with no “universal” background checks, gun homicides went from a slightly downward trend to a slightly upward one.

In all these changes, the degree of change was so slight as to be negligible. A 0.1 slope after passage in New York equates to 0.1 people per 100,000 citizens, or literally one-in-a-million (and for the observant among you, New York’s relative calmness balances out Delaware’s rise of slope from 0.1 to 0.2.

Much Ado About Nothing Much

We cannot say from this smallish sample of states that “universal” background checks do much of anything. The changes from before and after are basically flat, and with Connecticut we have a barrage of anti-crime initiatives and spending that make their law’s efficacy difficult/impossible to measure. Alone, these laws look like they produced no statistically valid change.

But measured against all states that had no “universal” background checks, there is a tiny amplifier. The difference between one-in-a-million people and two-in-a-million people is important only to that one extra dead person. In their fifth year of “universal”  background checks, long after any impact of the program would have been fully revealed, New York had 356 gun homicides. For most folks, the difference between 356 and 357 is statistical noise.