|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
GENDER DIFFERENCES AND PERCEIVED SANCTION THREATS: THE EFFECT
OF ARREST RATIOS
STEPHANIE E CARMICHAEL
A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF
FLORIDA INT PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF MASTER OF ARTS
UNIVERSITY OF FLORIDA
Stephanie E Carmichael
First and foremost, I would like to give my thanks and gratitude to Dr. Alex
Piquero, my committee chair, for the dedication he has shown me throughout the process
of completing this thesis and the many other proj ects that I have encountered since I
came to the University of Florida. I consider Dr. Piquero to be a role model and friend
who inspires me to always keep my goals high and my head up. I would also like to
thank Dr. Lonn Lanza-Kaduce, who provided much insight and guidance that really
helped to put this thesis in its current form. I feel extremely fortunate to be able to work
with these two extraordinary scholars and individuals.
I would also like to show my appreciation to other members of the Criminology
and Law Department who have supported me and my colleagues throughout the graduate
school experience. Their open doors and willingness to get involved have made all the
difference in the world.
Finally, I would like to thank my friends, both near and far, who have been the
most fun and greatest support group that a girl can have. At the top of that list is Steve
Miller, who has been by my side for six years of nonstop school and stress and has kept
me sane through all of it.
TABLE OF CONTENTS
ACKNOWLEDGMENT S .........__.. ..... .__. .............._ iii..
LIST OF TABLES ........._.___..... .__. ...............v....
AB STRAC T ................ .............. vi
1 INTRODUCTION ................. ...............1.......... ......
2 REVIEW OF LITERATURE ................. ...............3.......... .....
Deterrence Theory and Formal Sanctions .............. ...............3.....
Problems with Deterrence Theory Research .............. ...............7.....
Experiential Effects .............. ...............8.....
Gender and Deterrence ................. ...............11................
Current Focus ................. ...............16........... ....
3 METHODOLOGY ................. ...............18.......... .....
Data............... ..... ...............1
Dependent Variable ................. ...............19.......... .....
Independent Variables .............. ...............19....
Control Variables............... ...............2
Analytic Plan .............. ...............21....
4 RE SULT S .............. ...............22....
5 DISCUS SION AND CONCLUSIONS .............. ...............28....
LIST OF REFERENCES ................. ...............34................
BIOGRAPHICAL SKETCH .............. ...............39....
LIST OF TABLES
1 Experiential Effect-Full Model .............. ...............25....
2 Experiential Effect- Male Model ......... ........_____ ......... ...........2
3 Experiential Effect-Female Model .............. ...............26....
4 Formal Sanctions and Risk Perceptions-Full Model ................. .......__. ..........26
5 Formal Sanctions and Risk Perceptions-Male Model ................... ...............2
6 Formal Sanctions and Risk Perceptions-Female Model .............. ....................2
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts
GENDER DIFFERENCES AND PERCEIVED SANCTION THREATS: THE EFFECT
OF ARREST RATIOS
Chair: Dr. Alex Piquero
Major Department: Sociology
When studying perceptions of punishment certainty, researchers often focus on
either the effects of actively engaging in crime or the effects of formal sanctioning, and
occasionally both. Studies that have combined each of these angles have found that it is
important to assess arrest ratios when examining individuals' sanction perceptions.
However, studies that have examined arrest ratios and perceived sanctions have focused
only on male offenders. This paper furthers the research on this topic by examining a
sample of male and female Colorado inmates and measuring arrest ratios and sanction
perceptions for eight different crime types. OLS regression analyses reveal that gender
differences are evident among the sample, with males reporting significant experiential
effects for five out of the eight crimes as well as reporting arrest ratio effects for five of
the crimes. Females reported no significant experiential effects and female offenders in
only one crime type showed arrest ratio effects. Limitations and implications of this
research are then discussed.
Deterrence and rational choice literature has gone down many paths, first
measuring deterrent effects of sanctions at aggregate levels (Gibbs, 1968; Tittle, 1969;
Chiricos and Waldo, 1970), and then realizing that deterrence is a perceptual theory, at
the individual level (Waldo and Chiricos, 1972). Relating one's perception of sanctions
to their behavior is a key method of measurement used to support or refute the deterrence
doctrine as it relates to crime participation. Many researchers have used this same or
similar measure to capture how an individual's past behavior can relate to the current
perceptions and further how current perceptions affect future criminal behavior.
Once it was determined that micro-level research seemed to be the way forward,
deterrence research continued to branch out, encompassing research investigating formal
and informal sanctioning (Anderson et al., 1977; Green, 1989; Williams and Hawkins,
1986), certainty and severity (Waldo and Chiricos, 1972; Stafford et al., 1986; Grasmick
and Green, 1980), experiential and deterrent effects (Greenberg, 1981; Saltzman et al.,
1982), generalized others versus specific deterrence (Jensen et al., 1978; Paternoster et
al., 1983a; Silberman, 1976), as well as encompassing a multitude of different offenses
and deviant behaviors. Although the deterrence literature has been abundant, this body of
research has resulted in conflicting evidence depending on the data employed or the
behaviors measured (Nagin, 1998). These mixed results, along with additional gaps in
the research, make the deterrence doctrine and the rational choice framework a continued
interest in criminological work. One particular example of this kind of gap is that
researchers have been keenly interested in the experience of crime, the experience of
sanctioning, and the effects of gender on deterrence, yet no study has examined how
these key variables interact. Horney and Marshall (1992) elaborate on this issue:
The interpretation of the negative correlation between perceived risk and self-
reported criminality as a rational experiential effect rests on the implicit assumption
that people are generally successful in committing crimes. If, on the other hand,
people commit illegal acts but are caught and sanctioned, a rational choice
perspective would predict that the experiential effect would be modified; people
whose deviant behavior results in negative sanctions would (realistically) perceive
punishment as more certain than those who violate the law with impunity. It is thus
important to consider not only whether individuals engaged in criminal behavior,
but also whether they have experienced formal sanctions as a result of that
behavior. (p. 576-577)
With this line of reasoning in hand, several types of questions appear to guide the
path through the extant literature and the current focus of this study. This research is
guided by questions basic to the deterrence doctrine, including 1) does the experience in
formal sanctions have a deterrent effect, 2) is the experiential effect evident among a
group of offenders, 3) do perceptions of sanctions influence behavior, 4) is there a
relationship between the perceived risk of sanction, experience in crime, and the
experience of formal sanctions, and 5) what role does gender play within the deterrence
REVIEW OF LITERATURE
Deterrence Theory and Formal Sanctions
When deterrence theory emerged as an area of emphasis in the 1960s, mixed
reviews of its utility and validity were identified. Some of the early literature regarding
the deterrence doctrine found varied magnitudes of support for its basic tenet that
punishment, or the threat thereof, has an inverse relation with crime participation,
especially in regards to the certainty element of criminal deterrence (see Anderson et al.,
1977; Burkett and Jensen, 1975; Blumstein, Cohen, and Nagin, 1978; Cohen, 1978;
Geerken and Gove, 1975; Jensen, 1969; Jensen, Erickson, and Gibbs, 1978; Kraut, 1976;
Silberman, 1976; Tittle and Rowe, 1973; Waldo and Chiricos, 1972). More recent
research has shown support for the deterrence doctrine (see Nagin, 1998 for a review),
including research that has used a variety of approaches to measure the different aspects
of deterrence. For example, Stafford et al. (1986) found that a theory of deterrence fits
substantially better when certainty and severity measures are used interactively rather
than in an additive fashion. This is similar to the conclusion that Grasmick and Green
(1980) drew from their research in which they found that certainty and severity interacted
and could be operationalized as the perceived threat of legal punishment. Furthermore,
they concluded that perceived threats of legal sanctions, moral commitment to the law,
and social disapproval could be included into a simple additive model that reveals a
deterrent effect of the variables.
Additional research has also gone in many different, non-traditional directions,
including examining deterrence with different types of deviance such as shoplifting
(Kraut, 1976), cheating (Tittle and Rowe, 1973), intimate violence (Miller and lovanni,
1994; Miller and Simpson, 1991), traffic violations (Cohen, 1978), parking violations
(Chambliss, 1966) and drug use (Anderson et al., 1977; Burkett and Jensen, 1975; Waldo
and Chiricos, 1972; Smith and Paternoster, 1987), as well as utilizing different types of
samples. Some have aimed at expanding the utility and universe of the deterrence
doctrine, such as Piquero and Rengert (1999) who examined active residential burglars'
perceptions of risks and rewards through hypothetical scenarios of house burglaries.
They found that when burglars faced a hypothetical burglary situation with a high
probability of being caught, they were less likely to commit the crime, suggesting support
for deterrence. In another twist, Maxwell and Gray (2000) examined the perceived
certainty of sanctions among respondents who participated in an Intensive Supervision
Probation program in New Jersey. They found that there was a positive effect for the
probationer' s perceived certainty of sanctions and length of time staying in the program.
Some research has focused on whether the experience of formal sanctioning, as
opposed to perceptions of risk, does in fact indicate a deterrent effect. As with many
other aspects of the deterrence literature, results of these studies have also produced
mixed results. Some research has found evidence of a deterrent effect, including an early
(and likely methodologically problematic) study that found that for parking violations, an
increase in certainty and severity of punishment provided a deterrent effect (Chambliss,
1966). In addition, Kraut (1976) found that apprehension increased shoplifters' estimates
of formal risk, a finding similar to that of Paternoster and colleagues (1985) whose
analysis suggest, "that an increase in being formally sanctioned is related to an increase
in the perceived likelihood of one' s own punishment" (p. 426). Piliavin et al. (1986) also
found partial support that the number of prior arrests and experience in crime affected the
respondent' s risk perceptions, but this pattern was only seen in the youth portion of their
Nevertheless, it seems that much of the research has found little or no support for
the deterrent effects of formal sanctioning (Williams and Hawkins, 1986). Burkett and
Jensen (1975) found minor evidence that suggested being caught did not affect the
relationship between risk perceptions and self-reported behavior. Similar conclusions
were drawn by subsequent researchers, where Greenberg and colleagues (1979)
suggested that there was no deterrent effect for formal sanctions and two sets of studies
conducted by Richards and Tittle (1981, 1982) found no significant relationship between
arrest experience and risk estimates. This finding is also supported among a sample of
college students responding about drinking and driving when Lanza-Kaduce (1988)
found that the being stopped by police does not affect risk perceptions. Piquero and
Paternoster (1998) also considered drinking and driving and found that the experience of
arrest did not significantly affect projections to drink and drive. Apospori and colleagues
(1992) also found no support for a deterrent effect of sanctioning when examining a
group of respondents with varying levels of experience with the criminal justice system.
Although it seems that there is much evidence to suggest a lack of deterrent effect
of punishment, methodological issues might have lead to misleading results. One
criticism is that the samples of respondents employed often lack in the number of people
who have experienced sanctions (Horney and Marshall, 1992). Another issue is related
to the measurement of arrest as an absolute number. Richards and Tittle (1982) state
regarding their own research:
It would have been better if we could have employed a relative measure such as
number of times arrested relative to number of times that the individual violated the
law. This would have enabled us to more realistically tap the kind of experiences
that logically ought to influence perceptions about arrest chances. (p. 341, see also
Horney and Marshall, 1992)
This would support the use of arrest ratios over raw counts in order to assess how
experience in crime and arrest play a role in the deterrence process.
A third criticism deals with the measure of general sanctioning, rather than specific,
when interpreting the deterrent effect. This refers to measuring overall arrests, rather
than crime-specific arrest counts. This issue is identified by Horney and Marshall (1992):
If any arrest makes the risk of arrest for all crimes a more credible threat, then these
measures [general measures of formal sanctions] should be related to perceptions of
risk for each crime. If, however, an arrest for writing a bad check changes the
estimates of the likelihood of an arrest for writing a bad check but does not change
the perceived risk of using marijuana, predicted relationships may go undetected
when measures of general sanctions rather than crime-specific sanctions are used.
(p. 578, emphasis in original)
As a result of the combination of extremely varied results and methodological
issues, early deterrence research has left the state of deterrence theory a bit cloudy and
researchers have questioned the findings regarding the deterrence doctrine (see
Greenberg, 1981; Erickson, Gibbs, and Jensen, 1977; Minor and Harry, 1982, Saltzman
et al., 1982; Williams and Hawkins, 1986). In a comprehensive review of the state of
deterrence theory research, Nagin (1998) discusses the many different techniques used in
deterrence research and the limitations and gaps that extant research faces. Although
noting that research thus far is flawed, he concludes that the American justice system
does provide a deterrent effect on crime but that the current limitations make specific
deterrent predictions difficult.
Problems with Deterrence Theory Research
There has been a variety of research that finds that the deterrence doctrine is
lacking in several aspects. As early as 1972, some research had called into question
specific aspects of the deterrence doctrine, with Waldo and Chiricos (1972) finding that
even though perceptions of certainty impact participation in crime, the perceptions of
severity of punishment have no effect. Although Green (1989) did find a deterrent effect,
his study found that for drinking and driving, deterrence was only present when
considering informal and not formal sanctions.
In an even stronger argument against deterrence, Piliavin and colleagues (1986)
declared that deterrence and rational choice models of crime are too simplistic, and once
other factors (i.e. sex, previous crimes committed, etc.) are added into the model, neither
the perception of formal nor personal risks determined future behavior (see also Burkett
and Jensen, 1975). Some research has even suggested that there is no true deterrent
effect at all for formal sanctions (Richards and Tittle, 1981,1982; Greenberg et al., 1979).
In addition, there have been numerous studies which argue that a key issue that is
problematic in the early deterrence research is that the methodologies used in these
studies prevented researchers from ciphering out experiential effects from deterrent
effects. Here, researchers were concerned with the temporal ordering of the key deterrent
variables, realizing that the deterrent effects that emerged came from measuring the
inverse relationship between prior behavior and current perceptions. However,
researchers began to question whether the deterrent effect was actually being captured or
whether this type of research was actually revealing an experiential effect of being
involved in crime (see Greenberg, 1981; Saltzman et al., 1982; Paternoster et al., 1983a,
1983b; Minor and Harry, 1982). This issue has been further examined by numerous
researchers and found to be an important aspect of deterrence research. The next section
reviews the current state of literature regarding the role of experiential effects in
Although he never used the term 'experiential effect,' Greenberg (1981) clearly
addressed this issue in his response to Grasmick and Green's (1980) study that had
claimed to find a deterrent effect for three control-inhibitory variables on criminal
behavior. Greenberg argued that this claimed deterrent effect could be interpreted
differently, emphasizing the possibility that Grasmick and Green had reversed the causal
order of the variables. This meant that the variables (threat of legal punishment, social
disapproval, and moral commitment to the law), which Grasmick and Green claimed
inhibited participation in crime, could actually be the consequences of involvement or
non-involvement in crime. In order to help solve this problem, Greenberg suggested that
the use of longitudinal or panel data is needed and that cross-sectional data, like that used
by Grasmick and Green, are inadequate. Grasmick (1981) responded to these criticisms
by arguing that the data were not determining causality, but just inferring it based on a set
of assumptions and that the importance of the study was to bring together the refinement
of the measurements of deterrence and the confounding variables of various theories of
crime. Furthermore, he argued that the data currently available at the time did not allow
for researchers to conduct the ideal tests of deterrence and other theories of crime. Thus
what was important was to develop measurements and theory that can later be tested with
more adequate data.
One group of colleagues attempted to overcome the limitations of the early
deterrence studies by following Greenberg's (1981) suggestion and utilizing panel data
on the self-reported deviance of university students. Using longitudinal data, Saltzman et
al. (1982) found that the experiential effect is much stronger than the weak support found
for the deterrence doctrine, concluding that the "relationship between perceived sanctions
and behavior is likely reciprocal and processual" (Saltzman et al., 1982, p 184; see also
Paternoster et al., 1983a, 1983b). They found that since perceptions are not stable over
time, the experiential effect will be its strongest when the deviant behavior has just
occurred and that the effect of the behavior is strongest for first time users, suggesting a
novelty effect. In another study, these researchers further expanded upon these findings
by revisiting the issues of perceptual stability in samples of high school and college
students (Paternoster et al. 1983b). They found that for both groups, little perceptual
stability was evident, even within a six-month time span. Furthermore, they found that
the experiential effect was consistently stronger than the weak (although in some cases
significant) deterrent effect. They also found further support for their proposed novelty
effect, with considerable declines in the perception of risk for first time offenders from
Time 1 and Time 2. This same group of researchers further expanded on these initial
lines of reasoning, noting that the effects of perceived sanctions on criminal behavior was
minimal after controlling for factors such as moral commitment and informal sanctions
(Paternoster et. al, 1983a).
In a study that replicates the Eindings of Saltzman et al. (1982), Minor and Harry
(1982) found that perceptions of risk are not particularly stable and experiential effects
are consistently larger and more often significant than deterrent effects. Additionally,
Minor and Harry found that the experiential effect was strongest for those who originally
had higher risk perceptions (suggesting a naivete effect more than a novelty effect). In a
study that followed, Paternoster et al. (1985) tried to bring together the novelty effect
found in Saltzman et al. (1982) and the naivete effect from the Minor and Harry (1982)
study. The authors found that those who were inexperienced in committing the crime had
higher perceived certainty than the experienced offenders and that, in some cases, the
inexperienced reduced their perceptions more so than those who were less naive and less
inexperienced after committing a crime. They also found that for petty theft and bad-
check writing, a reduction in perceived certainty was significantly related to an increase
in the behavior. Finally, they found that an increase in formal sanctioning was related to
an increase in perceived certainty for the same two offenses.
Additional studies that examined the experiential effect have also been undertaken.
Miller and lovanni (1994) found support for an experiential effect for courtship violence,
with the prior use of violence diminishing the perceptions of risk for intimate violence.
Piquero and Paternoster (1998) tested Stafford and Warr' s reconceptualization of
deterrence and found that, as predicted, the more the respondents had successfully driven
while drunk, the more likely they were to report intentions to drink and drive in the
future. Apospori, Alpert, and Paternoster (1992) examined respondents with varying
levels of experience in the criminal justice system in order to assess the interactive effect
of experience in crime and prior involvement with the criminal justice system on sanction
perceptions. They found support for the existence of an experiential effect and
furthermore, found that those adults with more extensive experience in the criminal
justice sanctions reported an even stronger experiential effect than those with little or no
experience. Although this would be evidence contrary to deterrence, the authors
suggested that this pattern might have been observed simply because those with more
experience in the criminal justice system are offending at much higher levels than those
with little involvement (meaning that the experiential effect may not be working for those
with limited offending experience). Another explanation offered was that due to the
belief that one' s chance of being arrested is small, once they are arrested, their chances of
being arrested again will be reset and thus be unlikely (Pogarsky and Piquero, 2003).
Extant research has consistently found that the experiential effect is an important
issue and contribution to the deterrence perspective; however, much is left to be
explained. A review of the state of deterrence literature by Paternoster (1987) revealed
that issues of temporal ordering and spurious relationships tended to diminish support for
deterrence and iterated the importance of separating experiential from deterrent effects.
Additionally, the current review reveals a need to take the experiential effect one step
further to understand how arrest/formal sanctions operate within a rational choice model
that takes into account previous offending. In addition, gender issues must be taken into
account in order to fully understand deterrent and experiential impacts as well as further
understanding differences between male and female offending. The next section will
discuss what has been found so far regarding the role that gender plays in deterrence
perspectives and their experience of committing and sanctioning of crime.
Gender and Deterrence
Although largely ignored in the deterrence literature, the extant research that has
addressed gender has generally found it to be an important factor to consider with the
rational choice/deterrence perspectives. A large gap exists between the offense rates of
men and women, with little research explaining such a gap. In the literature that has
taken into account gender, much of it reported on differences between men and women
using gender only as a control variable (see Anderson et al., 1977; Jensen et al., 1978;
Silberman, 1976; Tittle and Rowe, 1973). Moreover, researchers were often focusing on
gender differences in delinquency (see Canter, 1982; Figueira-McDonough and Selo,
1980; Jensen and Eve, 1976; Smith, 1979), leaving much room to continue to explore
gender issues within deterrence. Even in the early literature, the role of gender in
delinquency and deterrence varied across different studies, with some research Einding
that gender had an important impact (Tittle and Rowe, 1973; Silberman, 1976) and other
research Einding that gender did not significantly contribute to deterrence components
(Anderson et al., 1977 ; Jensen et al., 1978).
Like many other aspects of the deterrence perspective, the role of gender within this
process is currently not clearly defined, with extant research finding mixed results on
how gender operates. When exploring marijuana use in males and females, Smith and
Paternoster (1987) stated, "It is currently unclear whether sanction threats are a more
effective deterrent for males or females" (p. 149). Although there is no agreement
throughout the literature, one Einding that has been replicated is that women tend to have
higher perceptions of sanction certainty than do men.
Richards and Tittle (1981) were among the first researchers to focus on gender
differences in perceptions of risk. They found that women had higher risk perceptions
than men for all the offenses they were queried about, but were quick to note that these
differences in perception could not account for the large disparity between men's and
women's offending. Their results suggested that differential visibility and differential
stakes in conformity were the most influential factors when explaining the differences
SAnderson et al. (1977) actually found that the perceptions of both formal and informal sanctions impacted
males more than females; however, the differences were so slight that they warrant a conclusion of "no
differences by sex" (p.109).
between men and women and concluded that sanction risks may be interpreted differently
depending on one's position in the social structure.
Additional studies have found evidence that supported the findings of Richards and
Tittle (1981). Finley and Grasmick (1985) explored differences in genders and
differences in traditional versus non-traditional females. They found that females in
general report higher levels of perceptions of legal sanctions than do males and
furthermore that traditional females seemed to be more susceptible to high perceptions of
punishments than were non-traditional females. Similar results were found when Tibbet
and Herz (1996) examined drunk driving and shoplifting in order to understand gender
differences in rational choice variables and the effect those variables have on intentions
to commit the offenses. First they found that nearly all of the rational choice variables
differed significantly between men and women, with men perceiving lower levels of
formal and informal sanctions. Second, they found that overall, the perceived threat of
shame accounted for most of the gender effect on offending intentions.
In a throng of research derived from the Oklahoma City surveys, a similar pattern
of gender and perceived threats emerged. Grasmick, Blackwell, and Bursik (1993 a)
looked at the relationship between gender and perceived threats of shame,
embarrassment, and legal sanctions for petty theft and assault (all of which could be
considered costs in the rational choice perspective). They first found that for all three
perceived threats, women scored significantly higher than men, with a small amount of
evidence suggesting that women and men were becoming more alike in their perceived
threats during the years between the two survey administrations. In another paper,
Grasmick, Blackwell, Bursik, and Mitchell (1993b) explored perceived threats of shame,
embarrassment, and legal sanctions in relation to interpersonal violence. Like before,
men perceived all three threats as lower than women but that for the threat of
embarrassment, men scored significantly higher in 1992 than in 1982. Blackwell and
Eschholz (2002) reviewed the same data and also found gender to be significantly related
to threat perceptions, but only found slight support for their prediction of gender
convergence. Finally, Blackwell (2000) employed one wave of this data in a test of
Power-Control Theory. She found a significant impact of gender on the perceived threats
of legal sanctions, shame, and embarrassment. In addition, she also found that for the
perceived threat of legal sanctions and embarrassment, these differences vary across
Despite the amount of research reporting gender disparities, several studies have
led to different Eindings. Smith and Paternoster (1987) tested the need and importance of
gender-specific theories of crime by studying male and female marijuana use. They
found that for this type of offense, traditional theories were not gender specific and
factors that influenced marijuana use were similar for males and females. They found
that the effect of the perceptions of certainty on marijuana use had an equally significant
negative effect for males and females, and concluded that, "It appears more likely that
sex differences in the volume of deviance reflect differential exposure to factors that
precipitate deviant behavior among both males and females" (p.156, emphasis in
original). Along these same lines, Piquero and Paternoster (1998) concluded from their
study of drinking and driving that there were many more similarities between men and
women than there were differences and that the same general form of the deterrence
model applied to both men and women. In addition, Uggen and Kruttschnitt (1998)
examined gender difference in crime cessation and concluded that regardless of gender,
rational choice variables were of little value in predicting desistance from crime.
However, they did note that women were more likely to leave their criminal ways behind
and remain crime free for longer periods of time, but they suggested that this could be
attributed to how men and women hold different places in social life.
Finally, two related studies did find gender differences, but in the opposite
direction of the previous studies. In the first, Miller and Simpson (1991) questioned
whether the factors that affect male perceptions of violence influenced females in a
similar manner. After analyzing data regarding courtship violence, they found that,
overall, men and women feel differential perceptions of sanction risks and severity, with
males perceiving more certainty in formal and informal sanctions. They also found
evidence for an experiential effect in males that affected perceptions of formal and
informal sanctions and for females in regards to attachment costs. In a similar study,
Miller and lovanni (1994) examined perceptions of risk regarding intimate violence and
found that females' perceptions of arrest certainty were actually lower than males. In
addition, they found that the effect of the prior use of violence (the experiential effect)
showed the largest difference between genders with regard to the perception of formal
risk and that males exhibited a significant experiential effect that was not found in
The extant research regarding gender and deterrence draw uncertain conclusions on
if and how gender impacts the deterrence/rational choice process. It seems that a pattern
has emerged suggesting gender is important to consider, and that females tend to perceive
higher sanction threats than males, although this was not the case across all studies. There
is also some evidence to suggest that gender is not an important factor in deterrence and
that in the case of rational choice variables, men are more similar to women. Overall, the
findings from the extant research leave much to understand about the impact of gender on
criminal involvement, including how experience in crime and arrest may operate
differently for males and females, especially any offender samples.
The point of departure for the current study is based on an earlier study that
examined perceptions of sanction certainty and involvement in crime and punishment for
a sample of male convicts. Horney and Marshall (1992) examined male, incarcerated
felons in order to assess the relationship among the perceived risk of sanctions, arrest
(sanction) history, and frequency of offending. The authors used a single-item measure
of perceived certainty of arrest for nine different crimes and calculated an arrest ratio for
the same crimes in order to determine the respondent's participation in crime. The first
key conclusion they drew was that the experiential effect (an increase in participation in
crime leads to a decrease in risk perception) is relevant for a sample of serious offenders,
suggesting generalizabilty of the findings. As a result, the authors concluded that it is
important to use a measure that takes into account the number of arrests related to the
number of crimes committed when studying sanction variables (i.e., arrest ratio). This
conclusion is drawn from findings which suggested that for eight out of the nine crimes
assessed, differences in risk perceptions of active offenders from non-active offenders in
that crime type were based on the ratio of arrests to offenses. Additionally, higher
perceptions of arrest were found for those active offenders who experienced larger arrest
ratios for four out of the six crimes that had at least 100 active offenders. The two
offenses that did not fit this pattern were assault and auto theft.
In sum, the current state of deterrence research leaves much to be understood, and
primarily how sanction perceptions are formed and modified in response to behavior and
its consequences (Nagin, 1998). The current study aims to contribute to the gap of
research that has left unexplained the extent to which the effects of experience in crime
and arrest are perceived differently by females and males. By using a sample that
includes both male and female felons, this research extends the current body of
knowledge by expanding that line of research into the role of gender. Based on the
review of the extant research, several predictions will be addressed in the current study.
It is hypothesized that 1) the experiential effect will result in lower perceptions of
certainty, 2) that arrest ratios will be important in that respondents who experience high
arrest ratios will provide higher certainty estimates than those with low arrest ratios, and
3) females and males will react differently to the experience of crime and arrest, with
females being more greatly impacted by these factors.
The data employed in this study come from self-enumerated questionnaires
completed by male and female inmates admitted into the Diagnostic Unit of the Colorado
Department of Corrections (English and Mande, 1996). Beginning in July 1988,
sampling was conducted one day a week through December 1989. Male respondents
were selected by allowing correctional officers to select inmates who had been admitted
within the last week from the most convenient cell block (inmates are randomly assigned
to the cells) and this sample represented anywhere from 25-100 percent of that week' s
inmate intake population. For females, every nth inmate from an alphabetical list was
selected during three separate visits to the Women's Facility and represented the entire
population of female prisoners rather than the intake cohort. The respondents were
surveyed in groups ranging between 10 to 20 inmates for the men and 35 to 45 for the
women. Although the inmates themselves completed questionnaires, help was provided
in calculating the number of street months and establishing a calendar for reference for
the crime participation. Additionally, pictures were used along side short descriptions in
order to assist the inmates in differentiating the crime types. The survey participation
rate was approximately 90 percent; however, an unknown number of inmates also refused
to leave their cells and never met with the researchers (although this is estimated to be
less than 10 percent). Participants were either given a long form (65 pages) or a short
form (45 pages) self-administered survey that was designed to assess criminal
involvement for the 12 months of street time (not incarcerated or hospitalized) prior to
their current arrest. The long form version will be used for the current study, which
includes 1146 respondents, with 91 percent male (1018) and 9 percent female (128).
Perceived certainty--the perception of sanction certainty was measured for each
crime that the respondent reported committing. Respondents recorded their perceptions of
the chance of being arrested when they were engaging in each of eight crimes (burglary,
robbery, assault, theft, motor vehicle theft, forgery, fraud and drug dealing) on a four-
point scale: 1-low, 2-some, 3-high, and 4-certain. This particular measure was used for
two reasons. First, crime-specific indicators were used for perception certainty in order
to prevent certain relations from going undetected through a more generalized measure
(Horney and Marshall, 1992), and second because perceived punishments for oneself
have been found to provide stronger relationships with self-reported criminality than
perceptions for a generalized other (Jensen et al., 1978; Paternoster et al., 1983a; Horney
and Marshall, 1992).
Gender--the sex of the respondent will be used in split-gender analyses for all the
models. The survey reports three measures of gender: anonymous male, confidential
male, and anonymous female. This will be dummy coded into two categories, 0=males
Arrest ratio--for each of eight crimes, respondents were asked questions
regarding their involvement in the crime and any experience of arrest for that crime
during the 12 street months prior to the arrest that led to the current incarceration. These
crimes were burglary, robbery, assault, theft, motor vehicle theft, forgery, and drug
dealing.3 Taking the count for the number of arrests in each crime and dividing it by the
number of times that same crime was committed determined the arrest ratio. This
measure was calculated for respondents who had been active in the specific crime during
the street months.
Following Horney and Marshall, there are several control variables included in the
models. The total number of crime-specific offenses and the total number of lifetime
arrests were included in the model as well as the respondent's age, race, level of
education, age at first arrest, and whether the respondent reported drinking alcohol
frequently during the street months. Due to the skewed nature of the number of crimes
committed variable, this variable was dichotomized by including those individuals
scoring under the 25th percentile (0, low number of crimes committed), and those scoring
above the 26th percentile (1, high number of crimes committed). The respondents' ages
ranged from 17 to 66, with a mean of 30.04 and a standard deviation of 8.21 years.
Almost forty-eight percent (47.8%) of respondents were white, 25.2 percent were black,
23 percent were Hispanic or Mexican, and 4 percent were of another race or the race was
unknown. The level of education ranged from 0 years to 18 years, with an average of
10.98 years and a standard deviation of2.05. The age at first arrest is used as a measure
2 This deviates from the Homey and Marshall (1992) study because street time was measured in their data
for a 36 month period.
3 This also deviates from the Homey and Marshall data. In the current data, counts for personal and
business robbery were combined into one question, although respondents were questioned on whether they
had or had not been involved with each. Homey and Marshall treat these as separate crimes.
of overall depth of criminal involvement (see Horney and Marshall, 1992) and had a
range of 6 years to 63 years, with a mean of 18.39 and a standard deviation of 6.77. To
assess if the respondents drank frequently during the street months, the question was
asked "During the street months, did you get drunk often?" Responses were labeled 'O'
for 'No' and '1' for 'Yes'. Overall, 62.4% of the sample responded that they did not get
drunk often and 37.6% reported that they did get drunk often.
The experiential effect, the effect of formal sanctioning, and a combination of the
two will be explored using split-gender analyses in order to assess gender differences.
The experiential effect will be assessed by using perceptions of risk as the dependent
variable and participation in each crime type, with age, race, level of education, age of
first arrest, and frequent alcohol use as independent variables. The next step will be to
determine the influence of formal sanctions on perceptions of risk. Once again a split-
gender analysis will be used to determine how the perceptions of sanction certainty for
each crime are a function of the arrest ratio, controlling for number of lifetime arrests,
age at first arrest, age, race, level of education, and frequent alcohol use. For all model
estimations, OLS regressions are employed because the outcome variables met OLS
The first set of analyses presented in Table 1 displays the experiential effect for the
full sample. Several findings can be noted from this table. First, significant experiential
effects can be observed for three crimes: burglary, assault, and motor vehicle theft. Two
other crime types, forgery and fraud, approach significance at p<. 10. The second Einding
is that the crimes display effects in the predicted direction, except for burglary. For all
crimes except burglary, perceptions of sanction certainty for each crime type are
decreasing as participation in that crime type increased during the twelve street months
confirming evidence of an experiential effect. The unexpected direction of burglary is
difficult to explain. When examining the mean amount of crimes committed, burglary
does not display a higher mean than the other crimes, indicating that the unusual finding
for the experiential effect of burglary is not due to excessive participation in that crime
compared to the others.4 This effect is also not due to extremely high estimates of
sanction certainty for burglary compared to the other crime types.' It may simply be that
burglary participants do not carry high certainty estimates to begin with and merely
increase their perceptions of certainty with participation and/or arrest, thinking that
sooner or later their luck will run out, a common finding in the qualitative literature on
property crimes (Tunnell, 1992). This point is returned to in the discussion section.
4 Mean number of times participating for each of the crimes: burglary (44.69), robbery (9.3 1), assault
(3.92), theft (41.45), motor vehicle theft (46.48), forgery (65.70), fraud (44.93) and drug dealing (495.46).
5 Mean estimates for perceptions of certainty: burglary (1.80), robbery (1.75), assault (1.72), theft (1.91),
motor vehicle theft (1.71), forgery (1.98), fraud (1.57) and drug dealing (1.83).
The next set of analyses examine if the experiential effect varies across genders for
a sample of convicted offenders. Table 2 presents the regression results for males for
each of the eight crime types and Table 3 presents these results for females for Hyve of the
crime types. The reason for the missing data in Table 3 is due to the lack of female
participation in burglary, motor vehicle theft, and fraud to adequately conduct regression
analyses. The most apparent Einding from these two tables is that the experiential effect
seems to be operative for males, but not for females. Males appear to be driving the
significant effects found for the full sample model, with significant experiential effects
for the crimes of burglary, assault, theft, motor vehicle theft, and fraud. In addition,
forgery approaches significance for males. Females show no significant experiential
effects, although they approach significance for robbery. It seems that there are apparent
gender differences for the experiential effect, with males' perceptions decreasing as
participation in the same crime increases and females showing no impact on perceptions
for crime participation. Despite the lack of statistical significance, the females do seem
to be coefficients that are quite large, in fact larger in some of the cases than the male
coefficients. This would suggest that the sample size may be an issue that hinders
statistical significance in the analyses for the female portion of the sample and that these
coefficients may provide significant results if the sample were larger. Finally, for all the
models that examine the experiential effect (full, male, and female) there are no distinct
patterns of significance that emerge for the control variables.
Tables 4 through 6 present the findings for the regression analyses that take into
account the influence of participation in crime and sanctions on perceptions of sanction
certainty. An arrest ratio was calculated for each participant for each crime type and is
included in the regression analyses to predict perceptions of sanction certainty. Table 4
presents the full model that includes both males and females. Five crimes are found to
have significant arrest ratio effects, including assault, theft, forgery, fraud, and drug
dealing. All the crime types with significant Eindings display the predicted direction of
the relationship, meaning that as the arrest ratio increases (increased number of arrests
related to number of crimes committed), perceptions of sanction certainty also increase.
Burglary and robbery, although not significant, display effects in the opposite direction as
was predicted, indicating that as the arrest ratio increases, perceptions of sanction
Tables 5 and 6 present the results from the regression analyses for the arrest ratios
for males and females, respectively. Once again, it seems that the males are driving the
Endings for the full sample model, displaying significant effects for the same crimes that
were significant in the full model: assault, theft, forgery, fraud, and drug dealing. In
practical terms, this suggests that males are updating their beliefs on sanction certainty as
a result of participation in these crimes and arrest. More specifically, as the number of
arrests increases relative to the number of crimes, males are increasing their perceptions
of sanction certainty. Females, however, do not seem to be affected by arrest and
participation in crime like their male counterparts, with the only crime that displays
significance for an arrest ratio being forgery. Table 6 presents the results for females.
Similar to the experiential effect analyses, the other variables that are controlled for in all
of the models display no discernable patterns that contribute to the overall outcomes.
Table 1. Experiential Effect-Full Model
Burlr Robbr Assault Theft MVT Fo er Fraud Drug Deal
n=211 N=108 n=257 n=194 n=104 n=148 n=63 n=267
B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta
248 138 130** -072 173 -041 -352 133 -172*** -123 146 -060 -447 176 -251* -301 206 -121 -440 274 -208 -049 121 -025
Age 1 st
010 016 046 018 019 103 008 013 043 018 013 115 001 020 002 -015 015 -105 009 021 064 003 012 018
Ae 011 011 080 008 013 061 021 010 145* 023 010 188* 012 015 089 021 015 149 020 016 187 007 009 048
Race -011 086 -009 049 127 039 -018 078 -015 018 087 015 -044 123 -035 234 129 152** -017 198 -011 -066 070 -059
Education -039 034 -083 011 047 024 -017 033 -032 017 035 037 069 041 178** 044 043 092 -042 051 -109 012 029 026
Drinkn 083 128 -046 031 177 018 026 130 013 -163 131 -088 -175 171 -100 -154 186 -069 202 254 109 -180 112 -107
*p<.05, **p<.10, ***p<.01
Table 2. Experiential Effect- Male Model
Burglary Robbery Assault Theft MVT Forgery Fraud Drug Deal
n=204 n=97 n=234 n=169 n=98 n=115 n=57 n=240
B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta
270 142 140** 014 180 008 -334 140 -163* -265 151 -135** -448 184 -254* -323 220 -139 -490 277 -245** -057 123 -030
Age 1 st
010 017 044 025 024 119 011 013 058 015 016 078 -009 021 -043 -027 019 -159 017 025 104 -005 012 -031
Ae 014 011 100 007 013 060 023 010 162* 024 010 191* 014 015 107 011 016 072 016 016 156 011 009 082
Race -014 089 -012 040 131 033 -001 082 -001 045 092 037 -001 127 -001 217 141 146 -041 201 -027 001 071 001
Education -042 035 -089 015 051 033 -011 034 -021 036 036 081 052 043 134 070 046 155 -067 054 -180 024 030 055
Drinking -088 131 -048 007 182 004 014 136 007 -053 137 -029 -134 177 -077 -245 202 -115 234 258 129 -157 113 -096
*p<.05, **p<.10, ***p<.01
Table 3. Experiential Effect-Female Model
B rlar Robber Asult T eft V F rer Fraud DrgDeal
n=7 N=11 n=23 n=25 n= =33 n=6 n=27
B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta
#of Crimes n/a -667 341 -416 -040 610 -020 569 579 248 n/a -838 643 -260 n/a -209 425 -080
Ae 1 st Arr -061 023 -649* -001 054 -004 032 042 335 -016 036 -146 053 033 324
Age ~010 035 -094 -029 052 -159 009 039 090 051 040 417 -052 042 -247
Race ~1 024 361 870* -036 296 -031 -163 253 -136 346 322 200 -891 268 -608***
Education 297 090 611* -134 170 -206 -110 148 -207 -102 123 -182 -132 102 -248
Drinkng -160 341 -100 211 500 101 -563 468 -268 172 469 068 -645 451 -264
*p<.05, *"p<.10, ***p<.01
Table 4. Formal Sanctions and Risk Perceptions-Full Model
Burlr Robber Asult Teft MV ogr raud DrqDeal
n=91 n=96 n=228 n=181 n=102 n=139 n=53 n=48
B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta
Arest Ratio -0642 086 -056 -0829 161 -057 398 107 244*** 259 075 244*** 129 164 080 612 250 206* 1 213 433 388*** 452 267 109**
Total Arrests 0020 005 036 0008 006 016 0045 005 059 0039 004 072 0075 011 088 0124 007 171** 0070 006 164 0000 004 000
Ae 1 st Arr -0045 019 -020 0142 022 083 0084 014 046 0121 014 082 0051 024 026 -0098 016 -069 0097 022 071 -0034 013 -020
Age 0043 012 028 0056 015 044 0224 011 160* 0221 010 182* 0150 016 109 0208 015 150 0034 020 029 0096 010 070
Race -0148 089 -013 -0282 138 -023 0142 080 011 0286 089 023 -0706 126 -057 262 130 167* 0637 213 040 -0737 073 -064
Education -0088 037 -019 0122 051 027 -0072 036 -013 0030 036 006 0475 042 125 0389 042 084 -0051 058 -013 0148 030 034
Drikin -0664 134 -038 0730 197 043 0578 136 028 -194 133 -106 -199 178 -116 -249 190 -111 164 278 089 -149 118 -087
*p<.05, **p<.10, ***p<.01
Table 5. Formal Sanctions and Risk Perceptions-Male Model
B rglary Robbery A sult T eft MV orgery Fraud DrgDeal
n=184 n=86 n=206 n=156 n=96 n=06 n=47 n=222
B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta
Arest Ratio 0700 087 -061 -104 164 -075 508 120 288*** 264 075 274*** 155 167 099 456 272 165** 1 302 418 450*** 439 260 115**
0016 005 029 0006 006 012 -0037 005 -049 -0040 004 -079 -0069 011 -083 0064 007 097 0071 006 179 0011 004 018
Age 1 st Arr -0061 020 -027 0208 029 100 0083 015 045 0069 016 038 -0057 025 -029 -0241 020 -141 0199 025 131 -0104 013 -062
Age 0076 013 050 0031 015 023 0245 011 178* 0218 011 171* 0177 016 133 0148 017 099 -0017 019 -015 0133 010 101
Race -0184 092 -016 -0427 146 -035 0058 083 005 0616 096 050 -0411 129 -034 257 146 171** 0509 216 033 0019 075 002
Eucation -0115 038 -025 0121 053 028 -0013 037 -002 0240 036 054 0254 044 066 0550 046 126 -0405 063 -106 0264 030 062
Drinking -0690 137 -039 0734 201 045 0395 140 019 -108 139 -061 -175 184 -104 -349 211 -163 183 284 103 -127 120 -078
*p<.05, **p<.10, ***p<.01
Table 6. Formal Sanctions and Risk Perceptions-Female Model
B rlar Robber A sult T ef V Forg r Fraud DrgDeal
n=7 n=10 n=22 n=25 n=6 n=33 n=6 n=26
B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta B Se Beta
Arrst n/a 382 464 217 -0460 321 -042 291 497 135 n/a 1 787 669 457* n/a 3 550 3 711 170
0794 067 796 -0180 036 -211 0082 032 077 0610 026 588* 0029 021 027
Ae 1 st Arr 0112 093 121 -0103 074 -054 0259 047 270 0182 044 168 0450 038 279
Ae -126 127 -1164 -0237 078 -130 0091 048 089 0015 049 012 -0495 047 -232
Race 1 660 645 1 426 204 349 157 -0921 265 -077 167 292 097 -820 295 -569*
Eucation 420 115 806** -180 207 -279 -145 162 -273 0399 107 071 -113 107 -213
Drinking -236 378 -139 440 551 206 -659 501 -314 0862 386 034 -616 460 -255
*p<.05, **p<.10, ***p<.01
DISCUSSION AND CONCLUSIONS
Horney and Marshall (1992) took an important step forward in deterrence research
by assessing risk perceptions among a sample of serious offenders. Their research filled
important gaps in the literature by first examining whether the experiential effect was
evident for different crime types among serious offenders as well as assessing how
formal sanctioning and the experience of crime together influence perceptions of
sanctions. The analyses conducted in the current research were meant to expand upon the
work of Horney and Marshall by examining these same issues on a sample of convicted
felons that included both males and females. Several hypotheses were outlined, with
partial and mixed support found.
First, it was hypothesized that the experiential effect would be evident among this
sample, indicating that lower perceptions of sanction certainty would result for those who
have increased involvement crime. Mixed results were found regarding this hypothesis,
with partial support resulting from significant negative experiential effects for the crimes
of assault and theft for the full model and for the crimes of assault, theft, motor vehicle
theft, and fraud for males. Females were not found to display significant experiential
Despite what seems as support for the first hypothesis, significant results were also
found for burglary, but in the opposite direction as expected. Reasons for this finding are
not quite clear and only speculations on why this might be the case can be made.
Burglars are often thought of as fitting subj ects for rational choice research since their
crimes are usually planned ahead, allowing for rational calculations of risks and rewards
(Piquero and Rengert, 1999). Extant research has found significant evidence that
burglars employ evaluations of risks and rewards; however the burglars often evaluated
the risks as low and chances of apprehension as slim (Bennett and Wright, 1984; Decker
et. al., 1993; Piquero and Rengert, 1999; Rengert and Wasilchick, 1985; Wright and
Decker, 1994). These findings may suggest one reason that the experiential effect was
found to work in the opposite direction for burglars. Perceptions of risk seem to be very
low to begin with for burglars, often so low that they believe that there is no chance of
being caught (Bennet and Wright, 1984). This could suggest that the perceptions of
sanction certainty can only change to reflect higher sanction risks as the burglars begin
and continue to offend, thus explaining the findings contradictory to what the experiential
effect would predict. Furthermore, if burglars do get apprehended, it may be that they
believe that their capture was due to the luck of the police or their own bad luck (Rengert
and Wasilchick, 1985), so their perceptions of risk may reset to the original low standards
they carried (see Apospori et al., 1992; Pogarsky and Piquero, 2003), thus explaining
why higher arrest ratios resulted in lower perceptions of sanction certainty. Still, this
interpretation is purely speculative and further research should be conducted to fully
understand the unique results for burglary.
The second hypothesis posed was that those offenders with higher arrest ratios will
provide higher certainty estimates than those who have low arrest ratios. Support was
found for this hypothesis, with five out of the eight crimes demonstrating this effect for
the full model and the male model. This would support the conclusion that Horney and
Marshall (1992:589) drew regarding sanction variables in that "it is necessary to use a
relative measure that takes into account the number of crimes being committed". Once
again, however, females did not demonstrate the same effects as their male counterparts,
with forgery serving as the only crime with an arrest ratio effect for females.
The third hypothesis predicted that males and females would display different
reactions to the participation in crime and arrest, thus resulting in greater changes in
perceptions of sanction risk as a result of increased sanction-to-number of crimes
committed for the females. Support was found for the first part of the hypothesis, that
differences do exist, however, it was not found for the predicted direction of the
difference. As previously stated, females do not seem to be influenced by participation in
crime or sanctions as do the males in the sample, with their sanction certainty perceptions
un-influenced by either of these variables. Findings from the split-gender analyses for
the experiential effect would suggest that gender is an important factor to consider in
these types of analyses, since males show significant relationships but females do not.
Findings from the split-gender analyses that included the arrest ratio produced similar
results as the experiential effect across gender. These findings may be attributed to
several reasons, including males' higher levels of participation in crimes, females' higher
sanction certainty estimates, or simply females being less affected by these experiences.
Naturally, there are limitations to this study. First, the sample size prohibited
performing regression analyses for several of the crimes for the females. Being able to
examine differences across all crime types could provide more insight into the differences
in perceptions of men and women, including comparing genders to better understand the
unusual direction of the experiential effect found by men for burglary. In the same vain,
even though only one of the female coefficients attained statistical significance, many of
the coefficient estimates are quite large--in many cases larger than the male estimates.
This raises another sample size issue, but also may indicate that with a larger n, the
female coefficients could become significant as well. Second, the respondents in this
study only answered questions regarding participation in the crimes for the twelve street
months prior to their current incarceration. This amount of time may be too short to
adequately determine the effect of crime participation. Although the significant results
would suggest that this measure is capturing the desired effects thus making it an
acceptable measure, different results might be found if a longer street time is taken into
consideration. The measures used here might also be reflecting a recency and primacy
effect that can diminish if both a longer street time were to be used or if the offenders
were no longer in prison.
Third, only formal sanction certainty was taken into account in the measures. It
may be that informal sanctions would demonstrate a larger effect on perceptions of
sanction certainty than do formal sanctions (see Paternoster et al., 1983a). Future
research should take informal and formal sanctions into account as well as other variables
that have been included in deterrence and rational choice literature. Additionally, the
variable assessing alcohol use during the street months could also be improved upon by
measuring whether alcohol and/or drugs were used during the commission of the act
instead of an overall measure of alcohol use. It is believed that the effects of alcohol use
may mediate the role that rational choice variables play during decision making (Assaad
and Exum, 2002).
Finally, the use of the arrest ratio may need to be further expanded. For example,
what does having a high arrest ratio really mean when linking it to perceptions of
sanction risk? As the results for Hyve of these crimes suggests, having a higher arrest ratio
leads to higher perceptions of sanction certainty. Does this suggest that deterrence is not
really working? If it were simply the experience of sanctions that drive up the
perceptions of sanction certainty and participation in crime was irrelevant, then the
deterrent effect would be working as predicted. However, if perceptions are affected by
experience in sanctions relevant to participation in crime, then this might suggest that the
deterrent effect does not emerge unless a person experiences an unusual amount of
sanctioning. This might suggest that there is a "tipping point" for deterrence and that
sanctions are ineffective until this point. Further research should delve into this topic
further to understand whether this phenomenon is evident and, if so, whether males and
females demonstrate different tipping points. This has obvious implication for deterrence
based public policies.
In the end, this research serves an important purpose within the deterrence literature
since it explores a topic that has been ill-studied. As gender considerations trickle into
research on crime, it becomes evident that the Hield does not have a firm understanding of
how gender relates to different aspects of the deterrence process. Differential patterns of
offending and reporting of perceptions by men and women is a topic that can still benefit
from future research. Continued research in this area should include gender as well as
further explore how perceptions are originally formed, if the formation process varies by
gender, and if the formation process determines how influential later experiences in crime
and arrest may be on perceptions. That several of the offenders in this study engage in
'belief updating' (Pogarsky and Piquero, 2003) suggests that sanction experiences do
serve to increase individual sanction certainty estimates, as deterrence would expect and
policymakers would hope for.
Learning behavior such as belief updating can have considerable implications for
deterrence based criminal justice policies and how the system handles offenders. If it is
true that some offenders are participating in belief updating, this shows that offenders are
susceptible to the effects of criminal sanctioning. More specifically, offenders are
learning what behaviors they can and cannot participate in without being apprehended
and can thus calculate a prediction of likelihood of being caught. This would suggest that
in order to increase offenders' perceptions of the risk of certain behaviors, it may be that
criminal justice policies need to place focus on increasing criminal justice contact with
offenders. Future research should continue to explore the impact of increased certainty of
sanction in comparison to the increased severity of sanctions to see which more greatly
impacts offenders' perceptions. If it is found that offenders perceptions are updated more
due to an increased certainty, less focus could be placed on severe sanctions and more
resources be placed on the detection and contact of current or suspected offenders.
LIST OF REFERENCES
Anderson, L.S., T.G. Chiricos, and G.P. Waldo. 1977. Formal and informal sanctions: A
comparison of deterrent effects. SocialProblems 25:103-114.
Apospori, E., G.P. Alpert, and R. Paternoster. 1992. The effect of involvement with the
criminal justice system: A neglected dimension of the relationship between
experience and perceptions. Justice Quarterly 9:379-392.
Assaad, J-M. and M.L. Exum. 2002. Understanding intoxicated violence from a rational
choice perspective. In A.R. Piquero and S.G. Tibbetts (Eds.), Rational Choice and'
Criminal Behavior: Recent Research and Future Challenges (pp. 65-84). New
York, NY: Taylor & Francis.
Bennett, T. and R. Wright. 1984. Burglars on Burglary. Hampshire, UK: Gower.
Blackwell, B. S. 2000. Perceived sanction threats, gender, and crime: A test and
elaboration of power-control theory. Criminology 38:439-488.
Blackwell, B. S., and S. Eschholz. 2002. Sex difference and rational choice: Traditional
tests and new directions. In A.R Piquero and S.G. Tibbetts (Eds.), Rational
Choice and' Criminal Behavior (pp. 109-136). New York, NY: Taylor & Francis.
Blumstein, A., J. Cohen, and D. Nagin. 1978. Deterrence and Incapacitation:
Estimating the E~ffects of Criminal Sanctions on Crime Rates. Washington DC:
National Academy of Sciences.
Burkett, S. and G.J. Jensen. 1975. Conventional ties, peer influence and the fear of
apprehension: A study of adolescent marijuana use. Sociological Quarterly
Canter, R.J. 1982. Family correlates of male and female delinquency. Criminology
Chambliss, W.J. 1966. The deterrence influence of punishment. Crime and'
Chiricos, T.G. and G.P. Waldo. 1970. Punishment and crime: An examination of
empirical evidence. SocialProblems 18:200-217.
Cohen, L. 1978. Sanction threats and violation behavior: An inquiry into perceptual
variation. In C. Well ford (Ed.), Quantitative Studies in Criminology. Beverly
Hills, CA: Sage Publications: Beverly Hills.
Decker, S., R. Wright, and R. Logie. 1993. Perceptual deterrence among active
residential burglars: A research note. Criminology 31:135-147.
English, K. and M.J. Mande. 1996. Measuring Crime Rates of Prisoners in Colorado,
1988-1989 [computer file]. Conducted by K. English and M.J. Mande, Colorado
Department of Public Safety, Division of Criminal Justice, Office of Research and
Statistics. ICPSR ed. Ann Arbor, MI: Inter-university Consortium for Political
and Social Research [producer and distributor].
Erickson, M.L., J.P. Gibbs, and G.F. Jensen. 1977. The deterrence doctrine and the
perceived certainty of legal punishment. American SociologicalReview 42:305-
Figueira-McDonough, J. and E. Selo. 1980. A reformulation of the "Equal Opportunity"
explanation of female delinquency. Crime andDelinquency 26:333-343.
Finley, N.J. and H.G. Grasmick. 1985. Gender roles and social control. Sociological
Geerken, M.R. and W.R. Gove. 1975. Deterrence: Some theoretical considerations.
Law and Society Review 9:498-5 13.
Gibbs, J.P. 1968. Crime, punishment, and deterrence. .Saida~ I'\ tei n Social Science
Grasmick, H.G. 1981. The strategy of deterrence research: A reply to Greenberg. Thze
Journal of Criminal Law and Criminology 72: 1 102- 1 108.
Grasmick, H.G., B.S. Blackwell, and R.J. Bursik. 1993a. Changes in the sex patterning
of perceived threats of sanctions. Law & Society Review 27:679-704.
Grasmick, H.G., B.S. Blackwell, R.J. Bursik, and S. Mitchell. 1993b. Changes in
perceived threats of shame, embarrassment, and legal sanctions for interpersonal
violence, 1982-1992. Violence and Victims 8:313-325.
Grasmick, H.G. and D.E. Green. 1980. Legal punishment, social disapproval and
internalization as the inhibitors of illegal behavior. Journal of Criminal Law and
Green, D. 1989. Measures of illegal behavior in individual-level deterrence research.
Journal ofResearch in Crime and Delinquency 26:253-275.
Greenberg, D.F. 1981. Methodological issues in survey research on the inhibition of
crime. Journal of Criminal Law and' Criminology 72: 1 094- 1 101.
Greenberg, D.F., R.C. Kessler, and C.H. Logan. 1979. A panel model of crime rates and
arrest rates. American Sociological Review 44:843-850.
Horney, J. and I.H. Marshall. 1992. Risk perceptions among serious offenders: The role
of crime and punishment. Criminology 30:575-594.
Jensen, G.F. 1969. "Crime doesn't pay": Correlates of a shared misunderstanding.
Social Problems 17: 189-201.
Jensen, G.F., M.L. Erickson, and J.P. Gibbs. 1978. Perceived risk of punishment and
self-reported delinquency. Social Forces 57:57-78.
Jensen, G.F. and R. Eve. 1976. Sex differences in delinquency: An examination of
popular sociological explanations. Criminology 13:427-448.
Kraut, R.E. 1976. Deterrent and definitional influences on shoplifting. Social Problems
Lanza-Kaduce, L. 1988. Perceptual deterrence and drinking and driving among college
students. Criminology 26:321-241.
Maxwell, S.R. and M.K. Gray. 2000. Deterrence: Testing the effects of perceived
sanction certainty on probation violations. Sociological Inquiry 70: 117-136.
Miller, S.L. and L. Iovanni. 1994. Determinants of perceived risk of formal sanction for
courtship violence. Justice Quarterly 11:281-312.
Miller, S.L. and S.S. Simpson. 1991. Courtship violence and social control: Does gender
matter? Law and Society Review 25:335-365.
Minor, W. and J.P. Harry. 1982. Deterrent and experiential effects in perceptual
deterrence research: a replication and extension. Journal ofResearch in Crime
and' Delinquency 19: 190-203 .
Nagin, D. S. 1998. Criminal deterrence research at the outset of the twenty-first
Century. In M. Tonry (ed.), Crime and Justice: A Review ofResearch, vol. 23.
Chicago: University of Chicago Press.
Paternoster, R. 1987. The deterrent effect of the perceived certainty and severity of
punishment: A review of the evidence and issues. Justice Quarterly 173-217.
Paternoster, R., L.E. Saltzman, G.P. Waldo, and T.G. Chiricos. 1983a. Perceived risk
and social control: Do sanctions really deter? Law and' Society Review 17:457-
Paternoster, R., L.E. Saltzman, G.P. Waldo, and T.G. Chiricos. 1983b. Estimating
perceptual stability and deterrent effects: The role of perceived legal punishment
in the inhibition of criminal involvement. Journal of Criminal Law and'
Paternoster, R., L.E. Saltzman, G.P. Waldo, and T.G. Chiricos. 1985. Assessments of
risk and behavioral experience: An exploratory study of change. Criminology
Piliavin, I., C. Thorton, R. Gartner, and R.L. Matsueda. 1986. Crime, deterrence, and
rational choice. American Sociological Review 51: 101-119.
Piquero, A.R. and R. Paternoster. 1998. An application of Stafford and Warr's
reconceptualization of deterrence to drinking and driving. Journal ofResearch in
Crime anadDelinquency 35:3-39.
Piquero, A. and G.F. Rengert. 1999. Studying deterrence with active residential
burglars. Justice Quarterly 16:451-471.
Pogarsky, G. and A.R. Piquero. 2003. Can punishment encourage offending?
Investigating the "resetting" effect. Journal of Research in Crime and'
Rengert, G. and J. Wasilchick. 1985. Suburban Burglary: A Time and a Place for
Everything. Springfield, 1L: Thomas.
Richards, P. and C.R. Tittle. 1981. Gender and perceived chances of arrest. Social
Forces 59: 1182-1199.
Richards, P. and C.R. Tittle. 1982. Socioeconomic status and perceptions of personal
arrest probabilities. Criminology 20:329-346.
Saltzman, L., R. Paternoster, G. Waldo, and T.G. Chiricos. 1982. Deterrent and
experiential effects: The problem of causal order in perceptual deterrence
research. Journal ofResearch in Crime anadDelinquency 19: 172-189.
Silberman, M. 1976. Toward a theory of criminal deterrence. American Sociological
Smith, D.A. 1979. Sex and deviance: An assessment of major sociological variables.
Sociological Quarterly 20: 183-195.
Smith, D.A. and R. Paternoster. 1987. The gender gap in theories of deviance: Issues
and evidence. Journal ofResearch in Crime and Delinquency 24: 140-172.
Stafford, M.C., L.N. Gray, B.A. Menke, and D.A. Ward. 1986. Modeling the deterrent
effects of punishment. SocialPsychology Quarterly 49:338-347.
Tibbet, S.G. and D.C. Herz. 1996. Gender differences in factors of social control and
rational choice. Deviant Behavior 17:183-208.
Tittle, C.R. 1969. Crime rates and legal sanctions. Social Problems 16:409-423.
Tittle, C.R. and A.R. Rowe. 1973. Moral appeal, sanction threats, and deviance: An
experimental test. Social Problems 20:488-498.
Tunnell, K.D. 1992. Choosing Crime: The Criminal Calculus of Property Offenders.
Chicago: Nelson-Hall Publishers.
Uggen, C. and C. Kruttschnitt. 1998. Crime in the breaking: Gender differences in
desistance. Law and Society Review 32:339-366.
Waldo, G.P. and T.G. Chiricos. 1972. Perceived penal sanction and self-reported
criminality: A neglected approach to deterrence research. Social Problems
Williams, K.R. and R. Hawkins. 1986. Perceptual research on general deterrence: A
critical review. Law and Society Review 20:545-572.
Wright, R. and S. Decker. 1994. Burglars on the Job: Streethife and Residential Break-
Ins. Boston: Northeastern University Press.
Stephanie Carmichael was born on October 31st, 1978, in Bowling Green,
Kentucky. After growing up in Fort Thomas, Kentucky, she returned to Bowling Green
to attend Western Kentucky University, where she earned her Bachelor of Arts degree in
psychology and sociology in 2002. She moved to Gainesville to enter the graduate
program in sociology at the University of Florida in fall of 2002, where she is pursuing
Stephanie currently has one journal publication forthcoming that looks at the effect
of anger on rational decision making, coauthored by Dr. Alex Piquero. She is currently
working on several other manuscripts in the areas of criminological theory and white
collar crime and will be continuing work on her Ph.D. in the Department of Criminology
and Law at the University of Florida.