This does not mean that mental illness is not a risk factor for violence. It is, but the risk is actually small. Only certain serious psychiatric illnesses are linked to an increased risk of violence.
One of the largest studies, the National Institute of Mental Healths Epidemiologic Catchment Area study, which followed nearly 18,000 subjects, found that the lifetime prevalence of violence among people with serious mental illness like schizophrenia and bipolar disorder was 16 percent, compared with 7 percent among people without any mental disorder. Anxiety disorders, in contrast, do not seem to increase the risk at all.
Alcohol and drug abuse are far more likely to result in violent behavior than mental illness by itself. In the National Institute of Mental Healths E.C.A. study, for example, people with no mental disorder who abused alcohol or drugs were nearly seven times as likely as those without substance abuse to commit violent acts.
Its possible that preventing people with schizophrenia, bipolar disorder and other serious mental illnesses from getting guns might decrease the risk of mass killings. Even the Supreme Court, which in 2008 strongly affirmed a broad right to bear arms, at the same time endorsed prohibitions on gun ownership by felons and the mentally ill.
But mass killings are very rare events, and because people with mental illness contribute so little to overall violence, these measures would have little impact on everyday firearm-related killings. Consider that between 2001 and 2010, there were nearly 120,000 gun-related homicides, according to the National Center for Health Statistics. Few were perpetrated by people with mental illness.