Research also has found that novice and experi-
enced officers employed different frames for making
decisions. In making arrest decisions, novice officers
focused on the blameworthiness of each party,
whereas experienced officers focused on their ability
to substantiate claims and the risk of future violence.
The shift from focusing on normative to efficiency
issues occurs relatively swiftly, after 1 year of service.
Officers also use stereotypes of domestic violence,
social class, mental illness, race, gender, and other
salient categories. Stereotypes help officers complete
missing information, interpret conflicting stories, and
make assumptions about likely outcomes or responses.
Research has found that experienced officers consid-
ered their stereotypic beliefs about battered women’s
propensity to use self-defense in arriving at their
arrest decision. Moreover, individual officers have dif-
ferent stereotypes about domestic violence, especially
regarding how much women provoke the violence or
react in self-defense. Officers inferred that men who
abused wives who were hallucinating or drunk were
less dangerous and that wives were more responsible
for the violence, suggesting that stereotypes about
mental illness also guide their interpretations when
suspects or victims exhibit mental illness. These
stereotypes thus affect officers’ inferences about the
situation and may lead officers to provide unequal
protection for victims who have a mental illness or
violate social or gender norms.Criteria Used in Arrest Decisions
Legal criteria that have been found to consistently
increase the likelihood of arrest include a disrespectful
attitude toward the police, the presence of witnesses,
the presence of a weapon, the presence of the perpetra-
tor, and a violation of an order of protection. Officers
typically make an arrest only in 20% to 50% of the
cases where there is clear evidence of a violation of an
order of protection. This finding indicates that officers
use their discretion and interpret the dangerousness
and risk to the victim in determining whether to make
an arrest when a perpetrator has violated an order of
protection. Several studies have found that a violation
of an order of protection increases the incidence of
arrest, but its effect on arrest is no greater than that of
other situational criteria.
Several criteria have been inconsistently related to
arrest decisions: the suspect’s gender or race, victim’s
or suspect’s use of alcohol, marital status of the suspectand the victim, presence of children, presence of
injuries, victim’s preference for arrest, and suspect’s
gender. The influence of these criteria depends on
other environmental and situational characteristics. For
example, research based on police reports found that
substance use or the presence of children decreased the
likelihood that the batterer would be arrested if the vic-
tim was African American but increased the chance of
arrest if the victim was Caucasian.
Research has found that arrest rates for cases
involving visible injuries vary from 30% to 73%
across police departments. Across archival and
vignette studies from the early 1980s until 2005, the
presence or seriousness of visible injuries is not suffi-
cient to invoke arrests, and its influence on arrest deci-
sions depends on other situational characteristics.
For example, visible injuries increased the chance
of arrest when the perpetrator is present but had no
effect when the perpetrator has fled the scene before
the police arrived. Officers also were more likely to
use the presence of visible injuries in their arrest deci-
sions when departments had a clear policy to arrest
when the victim has injuries or when the jurisdiction
had a coordinated response to domestic violence.
The importance of the victim’s preference in arrest
decisions clearly varies across departments, studies,
and cases. Police officers often do not include the vic-
tim’s preference in the police report even when it is a
standard part of the police form, which suggests that
it often is not an important criterion. Studies generally
find that the victim’s preference for arrest has a mod-
est impact, accounting for 4% to 5% of variation in
officers’ decisions on whether or not to make an
arrest. Officers often are not persuaded by the victim’s
preference for arrest because they think that most vic-
tims will drop charges, do not know what they want at
that time, or are not providing an honest account of
what happened. Officers’ stereotypes about battered
women and domestic violence also may affect how
they interpret the victim’s preference for arrest.
Police officers do not provide all battered women
with the same protection. Several studies have shown
that police officers are less likely to arrest perpetrators
who attack women who are drunk or having affairs.
Officers who use a normative frame are more likely to
arrest the husband if the battered wife is mentally ill
because they believe that he is more blameworthy for
hitting someone who cannot control her actions. In
contrast, officers using an efficiency frame are less
likely to make an arrest in this circumstance because568 ———Police Decision Making and Domestic ViolenceP-Cutler (Encyc)-45463.qxd 11/18/2007 12:43 PM Page 568