Journal of Criminal Law and Criminology

Volume 72
Article 15
Issue 2 Summer

Summer 1981

Victim-Offender Dynamics in Violent Crime

Richard Block

Follow this and additional works at: https://scholarlycommons.law.northwestern.edu/jclc

Part of the Criminal Law Commons, Criminology Commons, and the Criminology and Criminal
Justice Commons

Recommended Citation
Richard Block, Victim-Offender Dynamics in Violent Crime, 72 J. Crim. L. & Criminology 743 (1981)

This Symposium is brought to you for free and open access by Northwestern University School of Law Scholarly Commons. It has been accepted for
inclusion in Journal of Criminal Law and Criminology by an authorized editor of Northwestern University School of Law Scholarly Commons.
0091-4169/81/7202-0743
THEJOURNAL OF CRIMINAL LAw & CRIMINOLOGY Vol. 72, No. 2
Copyright 0 1981 by Northwestern University School of Law Pntedii US.A.

VICTIM-OFFENDER DYNAMICS IN
VIOLENT CRIME
RICHARD BLOCK*

A long held axiom of criminology is that in most societies crime

prevention is primarily the responsibility of the citizenry and not the
police. The study of victim-offender dynamics in violent crime is re-
search of events which result either from a failure to prevent crime or a
willingness to precipitate or participate in a criminal event. When pre-
vention has failed, the decision to resist and the method of defense be-
come important determinants of both the decision to invoke the
criminal justice process and the criminal processing system's decisions to
react to the crime. Thus, the perspective of the victim is probably more
important for understanding the violent criminal event than either that
of the police or courts.
The major responsibility for crime prevention has historically been
with the community and citizen, not with the crime processing system.
If the study of victims of crime is to have a major effect on rates of crime,
it may come through the enhancement of the ability of citizens to pre-
vent crime and react to criminal events in a way which minimizes the
resultant damage and injury to the victim.

I. THE CRIMINAL EVENT

The dynamic of victim-offender interaction is important to the un-

derstanding of the nature of violent crime, both for the outcome of the
crime itself and for the victim. It is also important for understanding
police and court decisions to apprehend and punish offenders. Violent
crime can be thought of as social behavior involving at least two actors
and their interaction. In any violent crime, there must be a target, an
offender, and their interaction. This triad can be called the crime event.
Like all social behavior, the crime event is surrounded by a history and
an environment which themselves alter victim, offender, and future
events.
The criminal event may be thought of as one instance surrounded
* Professor, Department of Sociology, Loyola University of Chicago.
 RICHARD BLOCK [Vol. 72

by a microenvironment of social relationships, physical structures, and

weapons of potential use, a macroenvironment of neighborhood and
community, a history of social relationships, and ideas of violence and
danger, self-defense, social class, and segregation. The two actors, vic-
tim and offender, interact with and are affected by these structures, but
they retain individuality; their behavior can never be fully predicted.
The outcome of the criminal event is often determined by the inter-
action of victim and offender. This interaction will in turn affect the
victim's decision to invoke the criminal justice system and the criminal
justice system's decisions to process and prosecute the crime event. Yet
both the criminal event and the criminal justice system are embedded
within the society and within their own macro and microenvironments.
Figure 1 illustrates these relationships. All elements of the society share a
common heritage and history. Important for the study of American vio-
lent crime are the traditions of frontier violence, racial segregation, arm-
ing of the populace, and many other factors.'

HISTORY

FIGURE 1
THE CRIMINAL EVENT

The macroenvironment of criminal behavior is the physical, social,

and economic structure of the community in which the crime occurs.
I See VIOLENCE IN AMERICA: HISTORICAL AND COMPARATIVE PERSPECTIVES (H. Gra-
ham & T. Gurr rev. ed. 1979) [hereinafter cited as VIOLENCE IN AMERICA].
 1981] VICTIMOLOGY SYMPOSIUM

Some communities are structured with a wide availability of targets for

criminal attack. There may be tourists in one neighborhood or many
homes which are unoccupied during the day in another. Other neigh-
borhoods may have fewer obvious targets. Some individuals have wide
opportunities for legitimate behavior; others have few. The
macroenvironment of the criminal justice system overlaps that of the
crime environment; however, the macroenvironment of criminal
processing can often be defined by structural capabilities and capaci-
ties-how many cases and what types the criminal justice system can
handle.
The microenvironment of criminal behavior is the immediate net-
work of events and structures surrounding the crime, such as the rela-
tionship of victim and offender, the location of a crime, and the weapons
available for use. The microenvironment of the crime processing system
is the characteristics of ,particular police, prosecutors, and courts at the
time of contact with the crime or criminal. As can be seen in Figure 1,
the micro and macroenvironment of a crime and the criminal justice
system overlap. Still, they are not congruent. Much of the environment
of crime is largely irrelevant to the criminal justice system, and the crim-
inal processing environment has little effect on criminal behavior.
If this model is correct, research which uses data gathered by the
criminal justice system to analyze victims may be distorted, as victim
surveys of the criminal justice system may be. The perspective of the
victim and offender is lost in the environment of the police. While the
victim may react to his treatment by police, he has little knowledge of
police work or its constraints.
Within its environment the criminal event is the initiator of out-
comes and actions. Thus, the interaction of victim and offender largely
determines whether the crime is a rape or merely an attempt and also
determines the level of injury in the crime. To a large extent the out-
come of a violent crime determines whether the police will be notified.
It is generally believed that the police are far more likely to be notified
of a homicide than an assault, and that they are more likely to be noti-
fied of a completed robbery than an attempt. Notification is the bridge
between the victim and the criminal justice system and between their
two environments. In notification, the two environments most clearly
overlap. Factors which influence the decision to inform the police of a
crime, such as degree of injury or outcome, are also factors which affect
the police decision to investigate.
Police and court action are clearly influenced by the crime event.
Most studies of police and court decisionmaking have found the nature
 RICHARD BLOCK [Vol. 72

2
of the crime to be an important and valid factor in decisionmaking.
Yet the criminal justice process is surrounded by a different environ-
ment than the criminal event, and decisions of the criminal processing
system are often wholly independent of the environment of the crime.
The concept of the crime event and its surrounding environment
and history will form a basis of this article and of the future research
proposed here. Thus, the crime event must be related both to its out-
come and to its surrounding environment. At the same time the envi-
ronment of the criminal act must be kept largely separate from that of
criminal processing.
Given the importance of the crime event for the victim of crime, it
is not surprising that much of the early research in victimology was con-
cerned with victim-offender interaction. The study of victims of crime
began with the study of victim-offender dynamics. Once the study of
victims went beyond descriptions of the spatial and demographic de-
scriptions of incidents, the first topics addressed were victim-offender re-
3
lationship and the victim as a generator of his own victimization.
Early research in victimology often concentrated on the degree to
which the victim could be considered responsible for his own victimiza-
tion. Mendelsohn developed a typology of six types of victims varying
from those who were more guilty than the offender and those who were
solely guilty to those who were guiltless. 4 Von Hentig also developed a
5
typology of victims and discussed victim-offender interaction as a duet.
Wolfgang and his students carried on this tradition in studies of various
6
categories of crimes.
Victim precipitation may be thought of as a failure of the social
control mechanism of crime prevention. Wolfgang and others have ar-
gued that much victim-precipitated homicide occurs in a subculture of
violence in which norms of interpersonal behavior are different from
those of the society as a whole. 7 Similarly, Amir, in classifying victim
precipitation of rape, judged behavior which was different from that
normally expected of women to be precipitating.8 On a broader level,
2 Ste, e.g., Black, The Social Organization of Arrest, 23 STAN. L. REv. 1087, 1107 (1971).
3See, e.g., H. VON HENTIG, THE CRIMINAL AND His VICTIM: STUDIES IN THE SOCIOBI-
OLOGY OF CRIME (1948); Mendelsohn, The Victimology, I ETUDES INTERNATIONALES DE
PSYCHO-SOCIOLOGIE CRIMINELLE 25 (1956).
4 See S. SCHAFER, VICTIMOLOGY: THE VICTIM AND His CRIMINAL 35-36 (1977) (ex-
plains Mendelsohn's six types of victims).
5 H. VON HENTIG, supra note 3, at 383-450 (ch. 12).
6 M. WOLFGANG, PATTERNS IN CRIMINAL HOMICIDE (1958); A. NORMANDEAU,
TRENDS AND PATTERNS IN CRIMES OF ROBBERY (1968) (Ph.D. dissertation, University of
Pennsylvania).
7 M. WOLFGANG & F. FERRACUTI, THE SUBCULTURE OF VIOLENCE (1967).
8 M. AMIR, PATTERNS IN FORCIBLE RAPE 259-76 (1971).
 1981] VICTIMOLOGY SYMPOSIUM

various scholars have argued that the high rates of violence in America,
when compared to other developed nations, result from a culture of vio-
lence. 9 A subculture of violence thus exists within the culture of vio-
lence.
Early studies of victims utilized records of the criminal justice sys-
tem. Thus, victim-precipitated homicides or rapes were those perceived
as victim-precipitated by the police, courts, medical examiner, or coro-
ner. This examination may be affected by the environmental perspec-
tive of the agency and only partially take into account that of the
victim.
These studies were greatly affected by the data source. Before a
crime becomes officially recorded, it must pass through several filters. If
these filters randomly select events, whether the interaction is studied
from the victim's perspective in a survey or the police perspective
through official records is of no great importance. However, the crimes
that are reported to the police are not a random sample of all crimes
which occur. First, the victim or observer must decide that the benefits
to reporting or the moral imperative are sufficiently great to require po-
lice notification. Then the police must decide that they have the re-
sources and interest to respond to the crime.
As will be shown, both victim and police decisions are affected by
victim-offender interaction. Crimes which are successful are more likely
to be reported to the police than attempts. The criminal justice system
records a far different set of crimes than do surveys of victims, and the
outcome of crime as reported in victim surveys is far different than that
reported in official records. Criminal justice system records are highly
appropriate to study the effect of victim characteristics or victim-of-
fender dynamics on police, prosecution, and court decisions. They are
not appropriate to study the background of the criminal events, the rela-
tionship of victim and offender, or the dynamic of victim-offender inter-
action. To study these, knowledge of the victim's perspective and
environment is most appropriate. These can only be collected through a
victim survey, but this must be a far different survey from those cur-
rently gathered.

II. DOES RESISTANCE AEFECT THE OUTCOME OF VIOLENT CRIME?

In a previous analysis of the character of robberies, aggravated as-

saults, and homicides in Chicago based upon records of the Chicago
Police Department, I found that death or injury and success or failure in
violent crime was to some extent determined by the nature of the vic-
9 See VIOLENCE IN AMERICA, ,upra note 1, at 17-18.
 RICHARD BLOCK [Vol. 72

tim-offender dynamic at the time of the crime's occurrence.' 0

In police records, I found that victim-offender dynamics in robbery
were strongly affected by the presence of a gun. In robberies with a gun
threat, force was much less likely to be used than in robberies without a
gun. The use of force was related to resistance by the victim, the suc-
cessful theft of property, and injury to the victim. I concluded from this
analysis that victim resistance only slightly reduced the probability that
the robbery would be successfully completed and greatly increased the
probability that the victim would be injured." Police records thus sug-
12
gested that victim resistance during a robbery made little sense.
Nevertheless, I was concerned that these relationships might be af-
fected by victim and police decisions to begin the criminal justice proc-
ess. However, I lacked the conceptual tools and data to test these
concerns. When the city tapes of the National Crime Survey became
available, I was able to consider these decisions to invoke the criminal
justice process.
This analysis found that victim resistance, completion of the rob-
bery, and police notification were strongly related in the victimization
survey. 13 Table I illustrates these relationships.

TABLE 1
PERCENTAGE OF ROBBERIES IN WHICH THE VICTIM
CLAIMS TO HAVE NOTIFIED THE POLICE,
CHICAGO, 1974
Completion Attempt
Resistance 66% (11716) 31% (17144)
No Resistance 57% (30057) 25% (3624)

This data indicated that attempted crime is less than half as likely
to result in police notification than completed crimes and that robberies
in which the victim resisted were slightly less likely to result in notifica-
tion than crimes with no resistance, independent of the effect of comple-
tion. Of the estimated 28,869 robberies in which the victim resisted,
40.6% were completed. Of the estimated 33,861 robberies with no resist-
ance, 90.1% were completed. Thus, robbery victims in the survey were
far more likely to resist than were victims in the police sample and that
resistance had a higher probability of success than in robberies recorded
10 The following section is a condensation and reinterpretation of Block & Block, Decisions
andData. TransformationofRobbery Incidents into OtwialRobbery Statistics, 71 J. CRIM. L. & C. 622
(1980).
11 R. BLOCK, VIOLENT CRIME: ENVIRONMENT, INTERACTION AND DEATH 87 (1977).
12 Id at 87-88.
13 Block & Block, supra note 10, at 636.
 1981] VICTIMOLOGY SYMPOSIUM

by the police. 4
The relationship between completion and notification is supported
by Hindelang and Gottfredson for the twenty-six-city sample. In these
surveys, 62% of completed robberies and 33% of attempts were reported
to the police.' 5 An analysis of the 1976 National Crime Survey (NCS)
results in almost the same conclusions as the Chicago analysis. See Ta-
ble 2. Resistance, completion, and police notification are clearly interre-
lated. The robberies that are reported to the police are more likely to be
successful, and the victim is less likely to have resisted than in those
robberies that ended without police notification.

TABLE 2
PERCENTAGE OF ROBBERIES IN WHICH THE VICTIM
CLAIMS TO HAVE NOTIFIED THE POLICE,
NATIONAL CRIME SURVEY, 1976

Completion Attempt
Resistance 60% (285) 32% (92)
No Resistance 61% (343) 36% (36)

Examining police records to determine the effect of victim resist-

ance results in a very different conclusion than from a victim survey.
Most of the cases of successful resistance will have been eliminated from
the data. Since cases of unsuccessful resistance are more likely to- be
recorded by the police, the researcher may erroneously conclude that
resistance is likely to be unsuccessful.
If one had sampled police-recorded robberies, it would appear that
victim resistance does very little good. Seventy-eight percent of resisted
robberies are completed, despite the resistance, and resistance only im-
proves the victim's chances of not having the robbery completed by
twenty percentage points. On the other hand, if one had used the vic-
tim survey as a sample, resistance would appear to be a more rational
act. The majority of resisted robberies are not completed. The differ-
6
ence is 48 percentage points.'
14 Id at 630 (Figure 5).
15 Hindelang & Gottfredson, The Victims DecisionNot to Invoke the Crimina/JusticeProcess, in
CRIMINAL JUSTICE AND THE VICTIM 66 (W. McDonald ed. 1976).
16 Block & Block, supra note 10, at 633 (Table A).
 RICHARD BLOCK [Vol. 72

TABLE 3
EFFECT OF SAMPLE ON CONCLUSIONS: ROBBERY
COMPLETIONS AND VICTIM RESISTANCE IN
THREE SAMPLES

Percentage
Victim Total Percent Points
Resistance Cases Completed Difference
Incident Yes 28,860 41%
Sample No 33,681 89% 48
Notified Yes 13,000 59%
Sample No 17,918 95% 36
Founded Yes 3,872 78%
Sample No 13,464 98% 20

Thus an incident sample, taken from a victim survey, describes the

victim often resisting, and then probably being successful in that resist-
ance. Police-recorded incidents, on the other hand, give an image of the
victim being less rational in resisting and more powerless to affect the
situation. Survey data describes a citizen who is more active in self-
protection.
The fact that characteristics of victim survey samples are systemati-
cally different from the characteristics of police data does not imply that
one is more accurate than the other. It only implies that the two are
measuring different phenomena.' 7 Police-recorded robberies may be an
appropriate data base for studying police activities or prosecutor and
court decisions. However, at least in Chicago, official records of robbery
paint a far different picture of the dynamics of victim-offender interac-
tion than do victim surveys. Given the major differences found when
samples of victimization survey robberies and robberies based upon po-
lice records are compared, several conclusions can be made about the
relationship between the crime event and its outcome.
In both the victim survey and police data, resistance is likely to
result in a reduction of the probability of completion of the crime. On
the other hand, in both the victim survey and police records, those who
resist are more likely to be injured than those who do not. However, as
Hindelang points out,' 8 a crucial cause and effect question remains.
Does resistance result from injury or lead to injury? While sequencing
of interaction was not possible with either sample, in both samples,
17 Id at 635-36.
18 M. HINDELANG, CRIMINAL VICTIMIZATION IN EIGHT AMERICAN CrrIEs 261 (1976).
 1981] VICTIMOLOGY SYMPOSIUM

physical resistance, the offender's use of force, and injury coincide. Eva-
sive resistance was less likely to coincide with force and injury.
Another paradoxical finding of this research and others1 9 is that the
offender's use of a gun threat reduces the probability that the victim will
be injured. The Chicago police data, however, indicate that it increases
the probability that the robbery victim will die. In the few reported
robberies in which a gun was used rather than threatened, 2.6% resulted
in death as compared to 0.6% of those robberies in which another form
20
of force was used.
Thus my research and that of others indicates that the dynamics of
the robbery event affect its outcome, the victim's decision to notify the
police, and early police decisions. However, the strength of these effects
are far different when victims are asked than when police records are
read.

III. VICTIM-OFFENDER DYNAMICS IN HOMICIDE

Most analysts hold that official records of homicide are far more
complete than those for other violent crimes. Thus, problems of sam-
pling are far less relevant than for robbery. However, while an incident
is likely to be recorded, the victim is dead. The nature of the victim-
offender dynamic in that incident must often be reconstructed by the
police. In this reconstruction, the victim's perspective is often repre-
sented by the offender or may remain unrepresented.
In my analysis of Chicago Police records I found it appropriate to
consider homicide generally to be the outcome of another violent crimi-
nal event (typically, either a robbery or an aggravated assault). A few
homicides could not be classified as the result of any other crime. How-
ever, when homicides were divided into those that resulted from argu-
ments or fights and those that resulted from robberies, many differences
could be found between the characteristics of the two forms of homicide,
but few differences between each type of killing and its precedent crime,
either demographically or in the relationship of victim and offender.
I named these two forms of homicide instrumental and impulsive.
Others have called them felony and nonfelony related. 21 In either case,
these two homicide forms are different not only as they occur, but also in
their legal implications.
If the identity of the offender in a felony-related or instrumental
19 Se, e.g., J. CONKLIN, ROBBERY AND THE CRIMINAL JUSTICE SYSTEM 120 (1972); J.
MACDONALD, ARMED ROBBERY, OFFENDERS AND THEIR VICTIMS 138 (1975).
20 R. BLOCK, supra note 11, at 58.
21 See Zimring, Eigen & O'Malley, Punishing Homicides in Philadelphia: Perspectives on the
Death Penalty, 43 U. CHI. L. REv. 227 (1976).
 RICHARD BLOCK [Vol. 72

homicide is known, the offender rarely is an acquaintance of the victim.

The offender is usually much younger than the victim, and the
probability of either victim participation or precipitation of the crimes is
small. Little can be said about resistance in instrumental homicide.
Many cases are not cleared and in many the nature of the victim's reac-
tion cannot be known.2 2 Impulsive homicides are far more likely to oc-
cur among relatives or acquaintances of nearly the same age, with a
higher probability of victim participation or precipitation. Impulsive
homicides are more likely than instrumental homicides to occur in the
home-hidden from public view.
The number of homicides increased very rapidly in most American
cities in the late sixties and early seventies. 23 In many cities rates of
homicide more than doubled. The increase continued through the mid
1970s and then stabilized. Although the motive for homicide varied
greatly from city to city in early studies,2 4 from Wolfgang's study of
Philadelphia in the 1950s on, all studies of homicide have found that
death resulting from an argument or fight was far more common than
from a felony. Still, much of the increase in homicides during the late
sixties and early seventies resulted from an increase in felony-related
killings.
An analysis of homicides in Chicago during the period from 1965
through 1976 indicates that both felony- and nonfelony-related homi-
cides increased rapidly fr6m 1965 to 1970. After 1970 assaultive homi-
cides remained constant or declined while robbery-related homicides,
especially with a gun, continued a linear increase through October 1974
and then rapidly declined. However, a more complete comparison of
the overall pattern of homicide with its different component patterns
indicates that all trends in homicide during the period 1965 through
1976 were accounted for by shifts in killings with a gun rather than by
shifts in the nature of the crime or changes in crime or community de-
mography. 25 Thus, it may be said that shifts in the number or propor-
tion of violent crimes in which a gun was used accounted for more of the
change in homicide number and patterns than did the dynamics of vic-
tim-offender interaction.
22 R. BLOCK, supra note 11, at 78.
23 See Barnett & Kleitman, Urban Violence and Risk to the Individual, 10 J. RESEARCH CRIME
& DELINQUENCY 111 (1973).
24 See Zahn, Homicide in the 19th Centug; United States, in HISTORY & CRIME: IMPLICA-
TIONS FOR CRIMINAL JUSTICE POLICY U. Inciardi & C. Faupel eds. 1980).
25 C. BLOCK & R. BLOCK, PATTERNS OF CHANGE IN CHICAGO HOMICIDE 26 (Ill. Law
Enforcement Comm'n 1980).
 1981] VICTIMOLOGY SYMPOSIUM

IV. VICTIM-OFFENDER INTERACTION IN AGGRAVATED ASSAULT

The problems of sampling are so severe in analysis of aggravated

assault that one might conclude that police reports of aggravated assault
are useful only because they represent the small set of incidents which
the police know about and are willing to retain on file. On the other
hand, victim reports in a survey may represent only those incidents
which the victim thinks are sufficiently impersonal to be reported to an
interviewer. The problems of accurate measurement of aggravated as-
sault began to surface with the beginnings of victim surveys. In 1966, as
scholars were obtaining the first returns from the NCS, it was clear that
some of the incidents reported as assaults were really fights and that the
designation of victim and offender was more determined by the sam-
pling frame than by any real difference in the behavior of victim and
offender. This realization plus a similar problem for fraud resulted in a
screen for all incidents to determine if they should be defined as criminal
acts.
At the inception of the NCS, the San Jose methods test indicated
that fewer than half the assaults reported to the police were also re-
ported to the survey interviewers. 26 A comparison of Uniform Crime
Reports for the twenty-six victim survey rates resulted in a negative cor-
relation. 27 Thus, it might be concluded that police records of assault
and victimization surveys are samples of almost wholly different phe-
nomena. My 1977 analysis of police-recorded aggravated assaults in
Chicago indicated that incidents called assault varied from husband-
28
wife domestic quarrels to crimes which looked very much like robbery.
Further analysis of the dynamics of victim-offender interaction through
these police records probably reveal as much about the police records as
about the nature of aggravated assault. Table 4 illustrates these rela-
tionships.
In police records most aggravated assaults were fights. Most of
these began mutually, and about 4% resulted in the death of at least one
-participant. A significant number, however, involved no victim partici-
pation or victim precipitation only as interveners. Crimes started by the
victim are more likely to result in the victim's death than others where
the nature of participation is known. However, dead men cannot tell
their own story and cannot say whether or not they began a fight. Over-
all, more than three-quarters of these police reports are records of crimes
26 National Inst. of Law Enforcement & Crim. Justice, San Jose Methods Test of Known
Crime Victims (Law Enforcement Assist. Admin. June, 1972) (Stat. Tech. Rep. No. 1).
27 Boland, Palternsof Urban Crime, in SAMPLE SURVEYS OF THE VIarIMs OF CRIME 31 (W.
Skogan ed. 1976).
28 R. BLOCK, supra note 11, at 89-90.
 RICHARD BLOCK [Vol. 72

in which two or more people actively participated either in a fight or as

interveners in a fight.

TABLE 4
VICTIM PARTICIPATION IN AGGRAVATED ASSAULT AND
OUTCOME, CHICAGO POLICE RECORDS, 1974

Victim Percent Percent Percent Percent Percent Percent

Participation Overall no injury some injury Death Guns Wounds
None 23 12 86 2 49 2
Fights
Off. Starts 3 7 92 2 15 29
Vic. Starts 6 4 76 30 48 23
Both Start 51 8 88 4 21 8
Intervent. 11 11 87 2 33 15
Unknown 7 9 75 16 27 3

Analysis of aggravated assault using current victim surveys is prob-

ably misleading. First, it is difficult to know how survey-reported as-
saults are related to the universe of all assault victimization and how the
effect of sampling affects the designation of an individual as a victim or
offender. Second, a large and unknown percentage of all assaults are
series victimizations by NCS definition and are neither included in NCS
publications, nor are they normally available in machine-readable form
for academic use.

V. PATTERNS OF VICTIM-OFFENDER DYNAMICS AND

OUTCOME IN RAPE

Growing awareness and concern for the victims of rape has resulted
in the creation of new programs to aid rape victims psychologically,
physically, and in their confrontation with the crime processing system.
One result of this increasing concern may be a shift in the probability
that a rape victim will notify the police of her assault. A second result
has been an increasing number of studies of the offense of rape, its vic-
tims and its offenders.
One of the earliest and most controversial studies was Amir's study
of Philadelphia based upon police records. 29 Several later studies based
upon police records failed to confirm Amir's findings. 30 However,
29 M. AMIR, supra note 8.
30 See, e.g., Chappell & Singer, Rape in New York City. A Study of Materialin the Police Files
and Its Meaning, and Chappell, Geis, Schafer & Siegel, A Comparative Study of Forcible Rape
Ofnses Known to the Police in Boston and Los Angeles, in FORCIBLE RAPE: THE CRIME, THE
VICTIM AND THE OFFENDER (D. Chappell, R. Geis & G. Geis eds. 1977).
 1981] VICTIMOLOGY SYMPOSIUM

Chappell concluded that police records of rape differed so greatly from

one American city to another that comparisons across cities were virtu-
ally impossible. In Amir's study of Philadelphia, 43% of offenders were
total strangers to their victims. In a study of Los Angeles, 56% were
strangers. In another of New York City, 72% were strangers. Finally, in
a study of Boston, 91% were strangers. 3 1 It may be that these relation-
ships do vary greatly between cities. Just as likely, however, is that po-
lice recording practices are not uniform.
While studies based upon police records probably reflect differences
in police practice as much as differences in the nature of rape, the use of
the NCS allows for a greater uniformity of reporting and more concern
for the victim's perspective. Griffin and Griffin have analyzed rape
from this perspective.3 2 They believe survey data may also have
problems, most notably underreporting of crimes which would reflect
poorly on the victim's own actions. Still, they conclude that victim-of-
fender interaction has an important effect on the probability of physical
injury. Using the NCS for 1973 and 1974, they found the offender's
threat to be a more important determinant of injury (Gamma=.56) and
completion (Gamma=.50) than resistance (Gamma=.36 and .31, respec-
tively). 33 Just as in my analysis of robbery, threat, resistance, injury and
completion interrelate. As in robbery, armed threat is not likely to re-
sult in injury. Griffin and Griffin believe that their findings suggest that
women may be well advised to resist their assailants with all means at
their disposal. They find that most victims (86%) will not sustain serious
34
physical injury whatever the resistance method employed.

TABLE 5
PERCENTAGE OF RAPES IN WHICH THE VICTIM CLAIMS TO
HAVE NOTIFIED THE POLICE, NATIONAL CRIME
SURVEY, 1976
Completion Attempt
Resistance 61% (23) 53% (66)
No Resistance 71% (7) 27% (11)

In Table 5, the analysis presented for robbery in Table 2 is repli-

cated for 1976 NCS rape victims. For these victims, completion, resist-
31 See FORCIBLE RAPE: THE CRIME, THE VICTIM AND THE OFFENDER, supra note 30, at
239, 259 (all four studies' results are presented).
32 B. Griffin & C. Griffin, Targeting and Aiding the Population at Risk (paper presented
at the Sw. Pol. Sci. Ass'n, Houston, Texas, 1978).
33 Id at 18.
34 Id at 19.
 RICHARD BLOCK [Vol. 72

ance, and police notification are also related. Attempted crimes are less
likely to result in notification than completed crimes, and crimes in
which there was resistance were more likely to be attempts than those
which were completed. The percentage of rapes in which the victim
reported resisting was far higher in 1976 (83%) than in Griffin and Grif-
fin's analysis of 1973-74 (56%). This may indicate a real and significant
change in women's behavior or a change in their description of crime.
In both surveys, a far higher percentage of rapes were attempts (72%)
than were robberies. Almost all robberies without resistance are com-
pleted regardless of what sample is analyzed.

VI. THE VICTIM IN THE CRIME PROCESSING SYSTEM

Very little research has studied the effect of the victim on the crime
processing system. Black has analyzed the relationship between the vic-
35
tim's desires and character and the police decisions to arrest a suspect.
Using data gathered while observing police in three cities, he concluded
that the victim's desires did make a difference. In about 40% of the
cases, the complainant did not clearly state a preference for further ac-
tion. In those cases where a preference was stated, 75% of the prefer-
ences were for police action. However, if the complainant wanted no
formal police action, the police always complied. When formal action
was requested, the police acted in 84% of the felonies and 64% of the
misdemeanors. Black saw this as an example of the conflict of universal-
ism and particularism in the criminal justice system. This problem he
believed was unsolvable except in a society in which all citizens agreed
36
on laws and enforcement.
While little can be known of the victim's desire in most homicides,
the decision to prosecute clearly hinges on the nature of victim-offender
interaction. Killings defined as justified are not prosecuted. However,
the definition of justification varies tremendously between jurisdictions.
In Lundsgaarde's study of Houston, 12% of all homicides were defined
as justifiable.3 7 In my research in Chicago in 1976, 3% were defined as
justified. 38 It would appear that the definition of justification is far
broader in Texas than in Illinois.
Recently, Williams has completed a study of the victim's role in the
prosecution of violent crimes using the PROMIS File for Washington,
D.C., in 1973. 39 She found that victim provocation as defined by the
35 Black, supra note 2, at 1095-96.
36 Id at 1106.
37 H. LUNDSGAARDE, MURDER IN SPACE CITY 237 (1977).
38 Unpublished data compiled by the author.
39 K. WILLIAMS, THE ROLE OF THE VICTIM IN THE PROSECUTION OF VIOLENT CRIMES
(Inst. for Law & Soc. Pol'y 1978).
 1981) VICTIMOLOGY SYMPOSIUM

prosecutor did affect the prosecutor's decisions. A personal crime which

included victim provocation or participation was half as likely to be
prosecuted as one which did not.
The analysis indicates that provocation or participation of the victim did
have an effect on the initial screening decision of the prosecutor but not on
the subsequent case processing decision. . . . [W]ith respect to the prose-
cutor's decision to dismiss a case after charges were filed and decision of
guilt made at trial, provocation was not a significant factor in any analy-
40
sis.
As Williams notes, these relationships are not unexpected. The
prosecutor will choose for further action those cases which he believes he
is most likely to win. Cases in which there is victim provocation or par-
ticipation are believed to have little chance for success.
Williams found that the relationship between victim and offender
as perceived by the prosecutor affected prosecution decisions for every
type of violent crime. She also found that many of the prosecutor's deci-
sions to drop a case resulted from the court's perception of complaining
witness problems in which victim and offender knew each other. How-
ever, she noted that much of this perceived noncooperation was based
on stereotypes or court errors rather than the witnesses' behavior.
Thus, limited research on victim-offender dynamics and the crimi-
nal processing system indicates: First, victim desires and behavior are
often taken into account by the police. Second, crimes in which there is
evidence of victim participation or provocation are not likely to be pros-
ecuted. Third, the relationship of victim and offender affects the prose-
cutor's decisions.

VII. A RESEARCH STRATEGY

This article has argued that the study of victimization must begin
with the world view of the victim. The particular concerns of the victim
and the macro and microenvironments that surround the victim are dif-
ferent from those of the police and the courts. Although these environ-
ments are interrelated, future research on victims of crime must have
more concern for the victim's perception of the criminal event than was
found in earlier research.
By concentrating on victimization rather than on victims, much
victimological research has been both trivial and expensive. The NCS,
by concentrating on the creation of police-independent, crime incident
estimates has eliminated much consideration of the nature of the phe-
nomena of victimization except for demographic correlates. Once these
few demographic correlates are analyzed, little else can be derived from
40 Id at 15.
 RICHARD BLOCK [Vol. 72

the victim survey. Furthermore, the noncrime variables included in the

NCS are mostly ascribed characteristics (such as age, race, and-sex)
which are by definition not manipulable through public policy. While
the NCS can give reasonably good overall estimates of victimization, it
is relatively useless in understanding the criminal event and its impact.
The NCS has been too concerned with estimation of crime rates
and the immediate economic and physical cost of crime and less con-
cerned with the background of these crimes, their outcome, and impact.
The need for estimation has led to data collection that could result in
research with little explanatory power.
On the other hand, studies of victims of crime based upon police
records, either as the framework for interview samples or as a represen-
tation of the reality of victimization, are also defective. The crimes
which become police records are not a random sample of all crimes.
They are a record of crimes in which police action was taken. Thus,
many crimes which were only attempts and many which the police were
too busy to handle are excluded from analysis. Police records are not
collected for criminological research. They are collected to fulfill the
efficiency goals of the department. As such, they represent a micro and
macroenvironment separate from that of the victim. Research on vic-
tims of crime has been largely barren, partly because either analysis was
limited to a few ascribed characteristics rather than the crime event, or
because the sampling frame was inappropriate for the study of victims
of crime.
The study of victims of crime must include not only the crime
event, but also its impact on the crime's outcome and the criminal jus-
tice system. It should include the study of the environment of crime, the
relationship of victim and offender, and their interaction. The criminal
event should also be related to the crime processing system through its
affect on the police, prosecutor, and court as they exist in their own
micro and macroenvironment.
In constructing a series of studies of victimization and the criminal
event from the victim's viewpoint, it should be noted that the study is
not only of the incidence of victimization, but also of primary crime
prevention. Most crime is prevented by its potential victims, not by the
police. The study of victims of crime is a study either of the failure or
impossibility of crime prevention. The study of victim-offender dynam-
ics in the crime event is the study of the actors' attempts to manage a
dangerous situation. Therefore, future research on victims should be a
series of interrelated studies including the following.
 1981] VICTMOLOGY SYMPOSIUM

A. A STUDY OF THE PROBABILITY OF VICTIMIZATION

This study would be similar to the NCS but with far greater elabo-
ration of behavioral and situational characteristics of victims and
nonvictims. In constructing this survey, far greater cognizance should
be taken of the value of the in-person interview for defining the respon-
dent's macroenvironment. Characteristics of the neighborhood, security
precautions taken, and home occupancy could be measured. 4 ' Recent
studies in the United States and the Netherlands 42 have emphasized the
importance of personal actions and security precautions as predictors of
victimization. Thus in this first study, the sampling framework of the
NCS would remain intact; however, additional information would be
gathered which would make better use of the in-person interviewing to
describe the environmental and behavioral characteristics of victims and
nonvictims.

B. A DETAILED STUDY OF THE CRIMINAL EVENT

The NCS has always been a very elaborate screening device for
detecting low incidence phenomena. The costs of screening are far
greater than the cost of interviewing victims. The second study would
be based on the NCS screen and would include all victims of very low
incidence phenomena, such as rape, and a sample of victims of higher
incidence phenomena, such as burglary or larceny. Questionnaires
would be designed separately for a description of the microenvironment
and victim-offender dynamics of each type of crime. There would be
seprate and distinct questionnaires for rape, assault, burglary, and so
on. In crimes of personal violence, victim-offender dynamics are more
important than in property crimes. Thus, interaction would be a far
more important component of the questionnaires for personal crimes
than for property crimes. While each of the samples would be represen-
tative of one sort of crime, the total of all samples could not be used to
depict crimes as does the current NCS.

C. A BRIDGE STUDY OF THE NOTIFICATION PROCESS

Included in each of the questionnaires proposed for study two

would be a far more specific and elaborate description of the notification
than is currently available. As previously discussed, notification is the
line between the criminal processing environment and the crime's envi-
ronment. Yet, the structure of the NCS notification questions make this

41 See Cohen & Felson, Social Change and Crime Rate Trends: A Routine Activity Approach, 44
AM. Soc. REv. 588 (1979).
42 J. VAN DIJK & C. STEINMETZ, DE CRIMINALITEITS OVERLAST VAN DE NEDERLANDSE

BURGERS IN DE PERIODE 1973-78, 29-32 (WODC Ministerie v Justitie 1980).

 RICHARD BLOCK [Vol. 72

bridge more like a short but dark tunnel. Each form of questionnaire in
Study B will be designed specifically to study the notification process of
that crime.
After the 1966 victimization studies, questions on the criminal
processing system from the victim's point of view were mostly aban-
doned. The victim may have little concept of the workings of the crimi-
nal processing system and its environments. The appearance of this
system to the victim may give insight into the citizen's concept on evalu-
ation and support of the police. Thus, questions should be included in
Study B which would describe the crime processing system from the vic-
tim's viewpoint-requests to sign complaints, police interviews, and
court processing. These should not be considered as a fully accurate
representation of the crime processing system, but only a description of
that system from the victim's viewpoint.
Also included in Study C would be the notification bridge between
the macroenvironment of the crime and the criminal process. This
bridge is a crucial link, yet little is understood about it. Once the police
are notified they are expected to act. A study similar to that of Black
and Reiss,4 3 but with a greater concentration on victim input might al-
low a description of the bridge process. However, the cost of a study
based on observation of all police squad activities might be very high.
The Reiss study required observation of police activities, very few of
which were related to serious crimes.
Thus the study of the bridge function from the police viewpoint
might be possible only through the reconstruction of those decisions us-
ing dispatch records as a sampling frame. This would of course elimi-
nate crime for which there was no dispatch record or for which no
formal police action was taken, but a street officer was involved.

D. THE VICTIM IN THE CRIME PROCESSING SYSTEM

Once a crime enters the crime processing system, the victim be-
comes only a minor actor. Williams has shown the possibility of analyz-
ing the victim's role as it is perceived by the system through the use of
PROMIS. 44 Other studies have confirmed this possibility. 4 5 The victim
is only one of many factors affecting the outcome of the crime as it
moves from police action through investigation, arrest, prosecution, and
conviction. It would be far more realistic to include an increased con-
43 Reiss, Public Perceptions and Recollections About Cn'me, Law Enforcement and CrznnalJustice,
in I STUDIES IN CRIME AND LAW ENFORCEMENT IN MAJOR METROPOLITAN AREAS (Pres.
Comm'n on Law Enforcement & Admin. of Just., Field Surveys III, 1967).
44 K. WILLIAMS, supra note 39.
45 See, e.g., J. EISENSTEIN & H. JACOB, FELONY JUSTICE: AN ORGANIZATIONAL ANALYSIS
OF CRIMINAL COURTS (1977).
 1981] VICTIMOLOGY SYMPOSIUM

cern with the role and character of the victim as an appendix to research
and analysis of criminal processing decisions than to field special studies.
Thus, Study D is not an independent study but a call for additional
concern for the victim in studying the crime processing system.

CONCLUSION

The study of victims of crime is most importantly the study of the

failure of crime prevention by citizenry and by the police and seconda-
rily the study of the active participation and precipitation of criminal
events by their victims. Early victimological research concentrated on
the active participation of victims of crime in their own misfortune. The
development of victimization surveys shifted the emphasis of research
away from the crime event to description of the overall incidence of
victimization and the incidence of victimization as defined for demo-
graphic subgroups of the population. The focus of victimization re-
search should shift from measuring the incidence of victimization to
defining the environment in which victimization occurs, from measuring
victim precipitation to describing the dynamics of interaction within the
crime event.
All crimes are events surrounded by a unique combination of micro
and macroenvironment and history. This article has summarized some
of the research on victim-offender dynamics within these unique struc-
tures. These dynamics may affect the outcome of criminal violence in
robbery, rape, and assault.
A series of four studies has been proposed to describe the role of the
victim by two methods. First, in the micro and macroenvironment of
the criminal event, an augmented victimization survey and crime-spe-
cific studies of victims should be used. Second, in the environment of
criminal processing a study of the notification bridge from crime to
criminal process and greater concern for the role of victims in research
criminal processing should be employed.
Part of the failure of victimology has been its failure to conceptual-
ize the role of the victim or to develop a unified body of research on the
victim's point of view. This article has proposed a method to unify re-
search and description of the victim of crime.