This article examines the extent to which arrest decisions in Montreal (199
1-1993) are significantly influenced by the racial status of defendants and
victims or by the racial composition of neighborhoods in which offending o
ccurs. Findings show that arrest outcomes for violent offenders are not aff
ected by the ethnic status of either defendants or victims. On the other ha
nd, the severity of arrest dispositions does vary across the urban landscap
e; specifically, it increases as the neighborhood residential concentration
of black communities increases. This mark-up in the severity of arrest dis
positions, however; affects both black and white defendants. Simply assumin
g that a significant proportion of policemen are "biased" against black def
endants does not account for the fact that arrest outcomes are exactly the
same for both the minority of black and the majority of white defendants li
ving in these mixed or heterogeneous neighborhoods. An alternative view-poi
nt could be that police officers expect or have learned to expect that pers
onal crimes or conflicts are more likely, in such neighborhoods, to produce
local disturbances.