It's pretty clear that owning a gun puts you at higher risk for harming yourself, but what about harming others?
A lot of the government-funded research on gun violence comes from the early to mid-1990s. That's because in 1996, the National Rifle Association successfully lobbied Congress to cut funding for gun violence studies. But before that, the CDC found that having a gun in the home made homicide about three times more likely for family members in that house [source: Sapien]. This jibes with a 1992 study, which found that family disputes that turned violent were three times more likely to result in death when a firearm was present versus other weapons [source: Saltzman].
Most homicides aren't carefully planned events. Instead, an argument with a friend or family member -- maybe over money or infidelity -- turns violent. Add a gun to the mix and the chances of death are greater than say, using a baseball or a knife.
But researchers point out that the correlation between homicide and firearms in the house is not as strong as the relationship between suicide and firearms in the house. Most murder victims are not shot at home – unless they are women, children or elderly. Also households with gun ownership might be more engaged in criminal activity. But one often-cited study of 400 homicide victims who were killed in their homes showed that half died of gunshot wounds and in the vast majority of cases, they knew the perpetrator. Forced entry was rare – only 14 percent of the time. Thirty-six percent of these homes owned firearms versus 23 percent of control households [source: Hemenway].
The change in behavior extends to the street. A 2009 study looked at 677 shootings in Philadelphia over two-and-a-half years and found that people who carried guns were 4.5 times more likely to be shot and 4.2 times more likely to be killed as unarmed people. The study authors think that guns may give their owners a sense of empowerment that leads them to act rashly or go into dangerous situations or places they might otherwise avoid [sources: Callaway,Branas].