By John M. Dawson and Patrick A. Langan, Ph.D., BJS Statisticians, 1994

This survey of murder cases that were disposed of in 1988 in the courts of 33 large urban counties reported:

Table 1 - Murder victims by victim-assailant family relationship, 1988

Relationship         Murder Victims
of Victim to     Number             Percent
to Assailant

All             8,063               100.0%

Nonfamily       6,755                83.8
Family          1,308                16.2

Spouse            528                 6.5
Offspring         285                 3.5
Parent            154                 1.9
Sibling           123                 1.5
Other             218                 2.5

Table 3. Alcohol use at the time of the murder, history of mental illness,
unemployment, and homelessness, by the family relationship of murder victims
and defendants, 1988

Relationship     Alcohol Use     History of
of Victim        at Time of      Mental         Unemployed     Homeless
to Assailant     the Murder(1)   Illness

All                 64.4%         4.3%             35.3%         1.6%

Nonfamily           68.0          2.7              36.6          1.7
Family              47.6         14.3              29.1          1.2

Spouse              54.4         12.3              25.0          1.6
Offspring           29.8         15.8              28.9            0
Parent              28.4         25.1              33.6          2.3
Sibling             53.9         17.3              34.9          3.3

(1) Alcohol use was coded only if alcohol was present in the defendant at the
time of the murder. This study did not report on drug use because of concerns
about the credibility of the drug use data.

According to these figures:

- A victim is 22 times more likely to be killed by a person without a mental
illness(MI) than with MI.

- A non-family member victim is 36 times more likely to be killed by a person
without MI than with MI.

- In every category, a defendant was more likely to be using alcohol than to
have a mental illness; overall, 15 times more likely. - In every category, a
defendant was more likely to be unemployed than to have a mental illness;
overall, 8 times more likely.


*If Dr. E. Fuller Torrey is correct that medication decreases violence, than
the 4.3% figure should be even lower today, due to improvements in medications
in the last eleven years.

* It's important to keep the proper perspective when interpreting these charts.
E.g., the 25.1% of parents killed by persons with a mental illness is 25.1% of
the 1.9% of parent victims, i.e. less than one half of one percent of all
homicides. It's also important to remember that this survey used a definition
of mental illness that included at least 25% of the population.