In 1998, Massachusetts passed what was hailed as the toughest gun-control legislation in the country. Among other stringencies, it banned semiautomatic “assault” weapons, imposed strict new licensing rules, prohibited anyone convicted of a violent crime or drug trafficking from ever carrying or owning a gun, and enacted severe penalties for storing guns unlocked.
... One of the state’s leading anti-gun activists, John Rosenthal of Stop Handgun Violence, joined the applause. “The new gun law,” he predicted, “will certainly prevent future gun violence and countless grief.”It didn't work out that way. While lawful gun licenses declined from 1.5 million in 1998, when the law was passed, to just 200,000 in 2002, Massachusetts went from 65 murders in 1998 to 122 in 1998. "Other crimes rose too. Between 1998 and 2011, robbery with firearms climbed 20.7 percent. Aggravated assaults jumped 26.7 percent." Gun control advocates' response to this is to blame other states; that is, they argue that absent all states enacting similar gun control laws, criminals will just cross borders and get guns from elsewhere. Jacoby goes on to question why, if this is so, no one mentioned it in 1998 when the bill was up for consideration. Read his whole article.
At a very gross, basic level, there is no correlation between the number of guns in citizens' hands and homicide rates. That is, ignoring all other factors, the percentage of firearms ownership cannot be used to predict a nation's homicide rate. For instance, according to a 2012 analysis undertaken by The Guardian newspaper the United States, which had by far the highest number of firearms per capita at 89 per 100 (which I believe to woefully undercount the number, which is probably more in the realm of 133 firearms per 100 people), but a homicide by firearm rate of only 3 per 100,000; as compared to Honduras, which has only 6.2 guns per 100 people but a homicide by firearm rate of 68 per 100,000.
A lot of anti-gun propagandists will refer to this as proof, so they say, that firearms do not reduce crime. In fact, the Real Clear Science article to which I link notes that if you eliminate undeveloped countries, and then ignore South Africa (which has an astronomical high number of homicides for a developed country), then there is a correlation between homicide rates and guns rates. That is, amongst developed countries, higher gun ownership is predictive of higher homicide rates ... as long as we ignore South Africa. Truly an example of the quip that if you torture the data enough, it will confess to anything.
But what does the data really show?
First, we have to ask why are the United States and South Africa such outliers? I would contend that it is because the United States and South Africa are the only developed countries that are truly multi-ethnic on a grand scale. As we know from looking at crime statistics from the United States, certain ethnicities are disproportionately responsible for homicides. In fact, if we were to remove homicides committed by "minorities," the homicide rates in the United States are comparable to those of most European countries (and I would note, even here, that it is still a relatively small number of individuals within a minority group that are responsible for the majority of the violence caused by that group). The same holds true internationally. The author of the Real Clear Science article decided to ignore undeveloped countries, but if we instead look at those countries, what we see is a strong correlation between high homicide rates and ethnicity, with the highest homicide rates in Latin America and Africa. So, what the international data is telling us is that putting firearms into the hands of a peaceful people does not turn them into bloodthirsty savages, and keeping firearms from a bloodthirsty people does not make them peaceful. Violent people will be violent and peaceable people will not, no matter what they are holding in their hands.
Second, the real issue really is not whether violent people with firearms will use those firearms to commit crimes, but whether peaceful people can use firearms to dissuade the violent people from committing crimes. We all know the answer to that: as John Lott's research has shown, and is implicit from Jacoby's article above, more guns in the hands of peaceable results in less violent crime. This is even implicit from the international data cited in the Real Clear Science.
Homogeneous countries offer poor evidence for gun crime because, as noted, a peaceable people will have lower crime rates regardless of the number of firearms; and a violent people will have high violent crime rates regardless of the number of firearms. What we have to look at are, in fact, the two outliers in the data--the United States and South Africa--because there you have a peaceful people attempting to live among a violent people. And focusing on those two countries, you see a dramatic difference.
Going into The Guardian's data set (you can download a spread sheet at the link), we see that South Africa has 12.7 firearms per 100 people, but a homicide rate (by firearms) of 17.03 per 100,000 and an overall homicide rate of 32 per 100,000. Conversely, as noted above, the United States has, according to The Guardian, 89 firearms per 100 people, and a homicide rate (by firearm) of 2.97 (and, as of 2016, an overall homicide rate of 5.3 per 100,000). So, in fact, the data suggests that a peaceful people enjoy a considerable benefit in possessing firearms when having to coexist with a more violent people.