Gun control proponents assume that clamping down on guns stops murderers. But does gun control work?

Over the past few decades, a variety of jurisdictions have brought in a wide range of gun control policies, reducing civilian access to firearms. Reviewing the research allows us to understand what has worked and what has not. We will focus exclusively on studies that use high-quality research methods that examine firearm homicide and total homicide rates. As Gary Kleck has observed in his award-winning book on gun control and subsequent articles, technically weak research mostly supports the assumption that gun control works, while stronger research does not.

Methodology is important. Improperly conducted studies have no evidentiary value. To assess the impact of legislation on homicide rates, a study must, at a bare minimum, compare rates both before and after any new law comes into effect. As surprising as it may seem, many claims of the effectiveness of gun control ignore this basic requirement. It is insufficient to claim, as many low-quality studies do, that variation in homicide rates across jurisdictions stems directly from differences in gun laws. To do so is to assume the conclusion. Crime rates may differ across nations (or regions) because of disparities in socioeconomic conditions or history. For example: The United States has a higher homicide rate than Canada for many reasons, including greater income inequality, catastrophic rates of violent crime linked to the drug trade, and the bitter heritage of slavery.

In addition to before-and-after comparisons, researchers must somehow account for the other variables that might affect homicide rate, such as economic cycles, changes in immigration patterns, and long term crime trends. This is the trickiest challenge for researchers as it involves complex statistical estimates.

Australia

Australia offers an excellent case study. Australia, like Canada, brought in stricter controls on firearms during the 1990s, banning self-loading firearms and pump-action shotguns. The 1996 firearms legislation sharply restricted legal ownership of firearms in Australia. It established a firearms registry, banning and confiscating over 650,000 firearms; it is estimated that over one million firearms were surrendered to police and destroyed.

Homicide rates (including firearm homicide rates) fell after the introduction of the firearms law. But, given that the homicide rates had already been falling prior to 1996, it is illogical to believe that this drop was due to the gun law.

To determine the impact of the 1996 law, five independent studies of the Australian firearm legislation have been conducted. Each of these studies not only did rigorous before-and-after comparisons, but they also conducted high quality time-series analyses, attempted limited multivariate analyses, and controlled for a variety of other influences such as police force and incarceration rates, and unemployment rate.

Not one of these studies could find evidence that the 1996 firearms legislation had a significant influence on the continuing decline in Australian firearm homicide rates, according to a recent review. Not one found statistical evidence of an actual impact of the legislative changes (McPhedran 2016). It is worth remembering a fundamental premise of scientific practice: the assumption that the null hypothesis is “true” until sufficient statistical evidence indicates otherwise.

Ignoring high-quality research, activists credit tighter gun law for the lack of mass shootings after 1996. This is unconvincing because there haven’t been any mass shootings in New Zealand either, which is quite similar demographically and economically to Australia, even though gun laws in New Zealand remain less restrictive than Australia, including allowing legal ownership of semi-automatic and pump-action firearms.

Absurdly, the 2014 Lindt Café attack in Sydney is used in support of the claim that Australian gun laws have been effective. It wasn’t a mass shooting but it could have been. A terrorist held 18 hostages using an illegally modified gun of a type already banned. Of course, the perpetrator had no shooter’s licence. Five hostages ended up being shot and two died.

Canada

In Canada, research by both Caillin Langmann and myself independently found no evidence that any of the Canadian gun laws were effective in reducing homicide rates. In 1992, I used a pooled cross-section, time-series model to evaluate the effect of the 1977 Canadian firearms legislation (Bill C-51) on the provincial homicide rate between 1969 and 1989. No significant effect of the firearms legislation was found on homicide rates. Several co-variates were included that were widely known as driving factors of homicide and violent crime: youth, immigration, population per police officers, the rate of prison incarceration, unemployment rate, percentage of the total population in the low-income bracket, and the Gini index of income inequality.

In 2012 Professor Langmann used three sophisticated statistical methods to search for any associated effects of firearms legislation. A wide range of important co-variates were included. No statistically significant beneficial associations between firearms legislation and homicide, firearms homicide, long-gun homicide, or spousal homicide rates after the passage of Bills C-51 (1977), C-17 (1991), and C-68 (1995). Nor were effects found after the implementation of firearms owner licensing in 2001 or the registration of rifles and shotguns in 2003.

A number of criminologists have reviewed the research on gun control and found it unconvincing.

The claim that “gun control works” contradicts the conclusions of the most respected researchers. There is no credible empirical support for restricting general access to firearms for civilians and expecting murder and other crime rates to fall. Exhaustive reviews of the gun control research by Gary Kleck, an award-winning criminologist, as well as two prestigious bodies in the US, the Centers for Disease Control and the National Research Council, have all independently concluded that empirical support for most common gun control measures in the United States is not convincing. Unfortunately, emotional issues are not settled with facts and this is something that continues to cost human life through bad legislation.

In a world dominated by Twitter comments, few read beyond headlines, often just leaping to conclusions. Because news reporters typically do not dig into details about research methods, the claims that are easiest to access on Google dominate the news cycle. Unsurprisingly, these are the most emotional and the least well researched. Even if readers wished to assess the methods used, it is frequently impossible to do because articles in medical or public health journals are often placed behind a pay wall.

Gun control supporters, such as public health activists, are quick to exploit every mass shooting. Their proposals for additional gun restrictions are often accompanied by studies claiming to be scientific, but minimal scrutiny soon cracks the thin guise of science.

Pseudo-scientific studies merely gull the unwary. Unlike criminology or economics, public health explicitly views its role as one of advocacy rather than discovery. Public health activists use the trappings of science to promote the pseudo-scientific paradigm that guns are a “disease vector.” A recent study published by the Bloomberg School of Public Health at Johns Hopkins University illustrates just how hollow are the claims typically made by public health activists. It is revealing that Bloomberg, a well-known financial backer of gun control organizations, contributed $300 million to his eponymous school in 2016.

One recent study exhibits the debilitating errors typical for public health researchers. The paper reviews an impressively large number of studies, but unfortunately the authors deliberately include studies with debilitating methodological errors. It is difficult to understand why any responsible researcher would include incompetent studies. The few high-quality studies are swamped by the larger number of incompetent ones.  Almost one third of the studies reviewed commit basic methodological errors that invalidate their results, such as not including before-and-after measurements, not including proper control groups, failing to control for potential covariates, or using inappropriate statistics, variables, or controls. Shockingly, much of the methodological details of this study is hidden behind a pay wall.

Conclusions

To sum up, there is no credible empirical support for expecting that murder and other violent crime rates would be reduced by laws that restrict general access to firearms for civilians. Unfortunately, straw men and red herrings have long dominated any emotional debate, and supporters of tighter gun control are quick to exploit every mass shooting. In a Twitter world, proposals for banning bump stocks, magazines, or even semi-automatic rifles that are black and look frightening to some, are featured prominently. The most strident claims are typically accompanied by studies claiming to be scientific. Later, with much less fanfare, minimal scrutiny uncovers the “scientific” claims as exaggerated or even bogus.

If increasingly restrictive gun laws can’t stop shootings by madmen or terrorist attacks, what could? An alternative to government would be to rely upon individual initiative and vigilance. This simply recognizes that solutions to problems are often best found closer at hand than the national government. Paper laws guarantee nothing.

One type of gun law  that does appear to be effective in reducing violent crime is allowing qualified residents who have been screened and do not have a criminal record to carry concealed handguns. In 1987 Florida was the first state to reform its law regulating concealed handgun carry, allowing qualifying residents to carry concealed handguns for the first time. This launched a wave of reform whereby in  2017 qualified applicants can legally carry concealed handguns in almost every jurisdiction of the United States.

The growing numbers of concealed handguns did not result in a blood bath. State records show that 6.5 % of American adults have concealed carry permits. Between 1991 and 2015, the US homicide rate fell at least as fast as homicide rates in countries that had tightened up their firearms laws. In the US, homicides fell 36%, while they dropped 21% in England, 20% in Australia and 19% in Canada over the same period.

A growing body of research shows that homicide rates fall after states allow residents to carry concealed handguns. John Lott’s findings are controversial and have stimulated other researchers to conduct their own analyses. After more than a decade of research, the dust is starting to settle: almost all published peer-reviewed studies find that right-to-carry laws in the United States reduce violent crime rates.

Solid research has also shown that the spread of concealed carry laws in the US has been one of the forces behind the declining American murder rates. Not the only cause to be sure, but one of them. Homicide rates have generally fallen in industrialized countries since the early 1990s, including in the US, but the US is unique in that while other nations were restricting civilian access to firearms, the US was liberalizing access to allow qualified civilians to carry concealed handguns.

In fact, just days before the terrible attack in Las Vegas, a concealed-carry permit holder stopped a mass shooting at a church in Nashville. Showing incredibly bravery, he physically attacked the shooter, wounded him in the melee, then retrieved his gun from his car, and held the injured attacker at bay until police arrived. The permit holder didn’t fire his gun, but he saved the lives of more than 40 people. Like other similar efforts by law-abiding firearms owners, his actions didn’t get receive much news coverage. Atrocities attract more eyeballs. And money.