Statistics can be pushed many ways depending on how you want to group the data (and they can all be perfectly correct all at the same time). Here's part of a paper that I wrote a while ago (maybe 2002?); so the data since then has changed some I'm sure.
"..Proponents are quick to point figures such as the ranking of the United States in firearm deaths. 14.05 out of 100,000 deaths are caused by firearms. Although a few other nations, such as Columbia (55.85 firearm deaths per 100,000 people) are worse off than the United States, there is a clear divide between the frequency of firearm deaths in the United States and other industrialized countries. Canada’s rate is 4.08 gun related deaths per 100,000 people, Germany’s: 1.47, the United Kingdom’s 0.48, France 0.67 (United Nations 106-107). Clearly we are at the head of the pack in this regard.
However, a breakdown of these figures shows some very interesting facets. While most people usually consider homicides to be the leading factor in these figures, reality lies elsewhere. The leading cause of firearm deaths in the United States is from suicide. Only 44% of gun deaths in the United States are from homicides. The majority of deaths are inflicted by the perpetrator on himself. Critics point out that if the guns aren’t available, the suicide rates would drop in step with the homicide rates; after all, about 63% of all suicides in the United States involve firearms (United Nations 55). Won’t people find other means to kill themselves though, if guns aren’t available? One may argue that the ease and quickness of using a gun increases the occurrence of unplanned suicides, where the victim shoots himself in a brief moment of despair. Studies have shown however, that the vast majority of suicides are planned out, with at least a day of preparation or deliberation (LaFollette 60). In addition, it is quite apparent that suicide rates are not related to gun availability or firearm homicides. Germany’s suicide rate, 15.8 deaths per 100,000 people, is over 10 times as large as its overall homicide rate, but less than 8% of suicides involved firearms. Japans suicide rate, 17.95 deaths per 100,000 people, one of the largest among in industrialized nations, dwarfs its firearm homicide rate, 0.03 deaths per 100,000 people, by a factor of almost 600! Yet only 0.22 percent of those suicides involved firearms. France and Canada, and Australia all have relatively high suicide rates (greater than 13 deaths per 100,000 people), yet the number of these deaths resulting from firearms is below 20% (in France’s case, below 5%). Clearly, gun availability does not affect suicide rates in any meaningful way. Just because over 60% of suicides in the United States use firearms, does not imply in any way that the overall suicide rate would change if the guns were not available.
Let’s for a moment ignore the factor of guns and just look at the homicide and suicide rates of different countries compared to ours. The rate of intended human deaths (suicides + homicides) in the United States would be about 17 deaths per 100,000 people. For comparison Germany’s would be 16.2 deaths per 100,000 people, Japan’s would be 17.97 per 100,000 people, Canada’s: 13.5 deaths per 100,000 people, France’s 18.1 deaths per 100,000 people, Switzerland’s: 21.8 deaths per 100,000 people, Finland’s rate: 28.1 deaths per 100,000 people, the United Kingdom’s: 7.68 (United Nations 108-109, 112-113). Does the fact that the majority of these deaths in the United States involve firearms make our problem any worse than any of these other nations? Although America is considered by many to be much more dangerous and violent than other industrialized nations, the truth is, we are right in the middle of the pack. Guns do not create the problem; they are merely used more often in the United States because of their availability.
Although our problem with homicides and suicides may not be as pronounced as some would make it out to be, the question at the core of the debate remains: would fewer guns in circulation in the United States lower overall crime rates? Historically, murder rates in the United States have varied quite extensively. Estimates of murder rates before the 1920s range from between 1-6 people per 100,000; however, there is wide consensus that the rate peaked in 1933 at about 10 per 100,000 people. The rate decreased progressively until just after World War II, when is rose slightly. It declined again until it hit a low point in 1958 at about 4.5 per 100,000 deaths. The murder rate remained steady until about 1965, where it began to rise rapidly. By 1972 the murder rate had risen to around 9/100,000. The rate remained virtually flat until 1982, when is declined marginally until 1986 where it rose again. By 1991, United States murder rate stood at 10/100,000. Soon thereafter the rate declined, this process has only accelerated with time. By 1998 the homicide rate stood at 6.3/100,000 down 30% from the 1991 rate (Vizzard 13). Explanations of this recent decline in murder rates overflow. Gun control advocates state the Brady law; on the other side, opponents credit the decline to relaxed restrictions on carrying concealed weapons. California politicians cite their three strikes policy. Police chiefs nationwide are quick to mention innovative police department strategies. It is of interest to note that this decline applied to almost every part of the country as well as Canada, hinting that a single factor would not explain the homicide rate reduction (Vizzard 16).
Anti- gun control lobbyists focus on two main reasons not to ban or further restrict firearms: self defense, and the second amendment. First let’s take a look at self defense. In 1975 a study by four Cleveland physicians concluded that the risk to a person housing a gun exceeded any benefit derived from the self defense provided (Cothran 34 ). The gist of this argument is that guns kept in the home are far more likely to be used in accidents or suicides rather than against crimes. The problem with this assessment remain quite apparent; while suicide statistics and firearm accidents numbers are reasonably reliable and easy to access, the data of guns used in self defense remains quite elusive. Attempts to estimate the number of incidents where firearms are used disrupt crime have yielded a wide range of results. Estimates of defensive gun uses (DGU) annually range from 50,000 up to 2.5 million annually. Given the wide range of estimates, it’s quite difficult to estimate firearms effectiveness in stopping violent acts. Using the figures given by the Department of Justice in its National Crime Victimization Survey (NCVS), we can place a rough estimate on the extents of DGUs. The survey estimates that between 64,000 and 85,000 DGUs take place per year in the United States. The survey found that 15.7 percent of the respondents reported that the DGU “almost certainly” prevented the attacker from killing them or any person; an additional 14.2 percent said the DGU “probably” saved their life or a life of another. Using the low end figure of 64,000, we can estimate that approximately 30% of DGUs almost certainly or at least probably save the life of the victim. That comes out to a bit over 19,000 lives possibly saved per year from defensive gun uses. With about 23,500 homicides per year in the United States, this number is quite significant (Vizzard 17-19). While gun control advocates respond to these figures, stating that respondents are quick to categorize the DGU as life threatening, when in fact, it isn’t. Even if this claim is partially true, it is quite apparent that DGUs play a significant role in interrupting violent crimes. Furthermore, most studies have a much higher estimation of the number of DGUs annually; it’s quite hard to ignore the fact that the 19,000 lives possibly saved per year could quite well be a low end figure..."