 

 

 

iziiiiiﬂjﬂiiiiiiiii’iii‘iiil

01581 5669

 

LIBRARY

Michigan State
University

 

 

 

Ilﬁsistocxnﬁﬂythatthe
dissertation entitled

Race, Urbanism, and Court Bureaucratization:
An Empirical Examination of Conflict-Weberian Theories

presented by

Florence Sylvia Ferguson

has been accepted towards fulﬁllment
of the requirements for

Doctor of ' degree in '

Doctoral Program with a Concentration in
Criminal Justice and Criminology

iplinary

Date December 5, 1996

MS U is an Aﬂ’trmaa’ve Action/Equal Opportunity Institution

 

 

042771

PLACE Ii RETURN BOX to remove thie checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU ie An Alﬁnnetive Action/Equel Opportmity inetituion
Wane-c

RACE, URBANISM, AND COURT BUREAUCRATIZATION:
AN EMPIRICAL EXAMINATION OF CONFLICT-

WEBERIAN THEORIES

By

Florence Sylvia Ferguson

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Criminal Justice

1 996

ABSTRACT

RACE. URBANISM, AND COURT BUREAUCRATIZATION:
AN EMPIRICAL EXAMINATION OF CONFLICT-
WEBERIAN THEORIES
By

Florence Sylvia Ferguson

Using a conﬂictNVeberian-based perspective, this study examined the
relationship between race, urbanism, court bureaucracy, and punishment. Conﬂict
theories argue that black offenders receive longer or harsher sentences than whites
because they are considered members of a subordinate population that is poor and
powerless. In contrast, the Weberian perspective views courts as bureaucracies; as
the size of the court increases, it becomes more bureaucratic, and efﬁciency
becomes the most important organizational goal. The purpose of this study was to
measure the effects of race, urbanization. and court bureaucratization on a sentence
of prison (IN) versus no prison (OUT).

Using a 1983 statewide sample of convicted felons from the State of
Pennsylvania, the analysis for this study proceeded from bivariate correlations to
logistic regression equations. Race did inﬂuence the sentence outcome; however.
racial disparity was found among male offenders. not females. These results

support conﬂict theory because they revealed black males were more likely to be

Florence Sylvia Ferguson
incarcerated than all other offenders. More important, the interactive effects
between gender, age. and court size reject the major tenets of Weberian theory
because the ﬁndings of this study indicate that older male offenders, particularly
blacks. were more likely to be incarcerated, even when they were sentenced in
urban court jurisdictions.

Finally, this study also revealed extralegal and legal variables were better
predictors of the INIOUT decision than contextual variables. The strong bivariate
correlations among the contextual variables made it difficult to determine the extent
to which they may have inﬂuenced the sentence outcome. Because researchers
have recently used contextual variables to explain sentencing practices, it may be

too early to conclude whether they are useful in predicting sentence outcomes.

Copyright by
Florence Sylvia Ferguson
1 996

ACKNOWLEDGMENTS

My ﬁrst acknowledgment is to God. Writing this dissertation has been a
challenging task, however one that i did accomplish. My second acknowledgment
goes to my loving father, Leon John, and to the memories of my mother, Sylvia, and
my young brother, Roger, who both are now resting in peace. I have ﬁnally
completed the task Mommy asked of me; now I can move on with my life. I truly
appreciate the continued support I received from Daddy throughout this ordeal. He
was not only ﬁnancially supportive, his love and conﬁdence that i would complete
this task never failed. I also would like to acknowledge all of my siblings (Barbara,
Leon, Jr., Archie, Johnny. and Rodney), especially my oldest sister. JoAnn, whose
continued emotional and loving support helped me complete this work. Next, I must
not forget my friends who gave me support. They include Roz, Marilyn and Roy.
Teresa, Gail, Saundra, Kim, and Ken. Most of them knew the problems I incurred
during this lengthy process; however, they gave me their support and the assurance
it would soon be over. Most important, I must mention Joshua Bagakas. who
worked with me over the past 10 years of this process. Had it not been for Joshua.
I probably would have never completed this work. His patience and guidance
cradled me through this process. He was so calm, when I was so upset. He gave

me what I needed to cross the ﬁnal 't" and dot my last sentence with a period.

V

Thanks are also in order for my guidance and dissertation committees. I
would like to thank Dr. John McNamara, who during the ﬁrst draft of my work retired
for medical reasons. I appreciate his conﬁdence in me, and although he was not
with me in the end, I will never forget the help he provided me in the early stages.
Next, I would like to give my sincere thanks and express my sincere gratitude and
appreciation to Dr. Harold Spaeth and Dr. J. Kevin Ford. These two guidance
committee members hung in there with me for the past 12 years; I thank them for
never abandoning me during this process. It is for these reasons I am glad i chose
them to guide me through this work. Finally. I would like to express my sincere
thanks and gratitude to Dr. Merry Morash, my chair. for her directions and for helping
me to ﬁnally bring closure to this dissertation. Dr. Charles Corley and Hazel Hardin
should also be mentioned and thanked for their continued support. I can’t believe

it’s over!

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................ ix
Chapter

I. INTRODUCTION ...................................... 1
ll. THEORETICAL PERSPECTIVES ........................ 11
The Consensus Perspective ............................ 11

Conﬂict Theory and the Threat of the Subordinate
Population ........................................ 12

The Weberian Theoretical Perspective and

Bureaucratic Justice ................................. 16
Integrating Theoretical Perspectives ...................... 22
Modeling the Sentencing Decision ....................... 27
Hypotheses ......................................... 28
Extralegal Variables .................................. 30
Race ............................................. 30
Gender and Age .................................... 32
Legally Relevant Variables ............................. 35
Prior Record ....................................... 35
Severity of Convicted Offense (Severity) ................. 36
Number of Current Convictions (Convictions) ............. 37
Contextual Variables .................................. 37
Urbanization and Structural Variables ..................... 39
Size of Subordinate Population ........................ 41
Unemployment ..................................... 43
Crime Rates ....................................... 46
Court Bureaucratization .............................. 50
III. METHODOLOGY .................................... 54
Data ............................................... 54
Variables and Measurement ............................ 55
Dependent Variable ................................. 55

vii

Independent Variables ............................... 57

Data-Analysis Procedures .............................. 59
IV. PRESENTATION AND ANALYSIS OF FINDINGS ........... 63
Descriptive Statistics for the State of Pennsylvania and
Urban, Suburban, and Rural Court Jurisdictions ........... 63
Descriptive Statistics for Extralegal, Legal, and
Contextual Variables ................................ 66
Relationship Between Extralegal Variables and the INIOUT
Decision .......................................... 68
Relationships Between Contextual Variables and the INIOUT
Decision .......................................... 70

Results of the Logistic Regression Model on the Effects of
Extralegal, Legal, and Contextual Variables and the

INIOUT Decision ................................... 73

V. SUMMARY, DISCUSSION, CONCLUSIONS, AND
RECOMMENDATIONS ................................ 76
Summary and Discussion .............................. 76
Conclusion ......................................... 82
Recommendations for Future Studies ..................... 85
REFERENCES ................................................. 96

viii

LIST OF TABLES

Table Page
1. Description of Variables, Coding, and Summary Measures ......... 56

2. Descriptive Statistics for the State of Pennsylvania and Urban.
Suburban. and Rural Court Jurisdictions ........................ 64

3. Descriptive Statistics for Extralegal, Legal, and Contextual
Variables and the INIOUT Decision ............................ 67

4. Correlation Matrix for Extralegal, Legal, and Contextual
Variables and the INIOUT Decision ............................ 69

5. Results of Logistic Regression Model for Extralegal, Legal, and
Contextual Variables and Interaction Terms for the INIOUT
Decision ................................................. 73

CHAPTER I

INTRODUCTION

The issue of racial discrimination in the criminal justice system has always
been of concern to social scientists. The claim of racial and class discrimination is
an essential proposition in recent theories on administration of justice (modern
conﬂicttheories, radical criminology, social justice perspectives, and labeling theory).
Unfortunately, the intensity of that concern has not always been translated into
effective and reliable investigative strategies. Although most of us have been led to
suspect that discriminatory practices exist, research in this area has yielded
inconsistent and even contradictory ﬁndings. Several major reviews of this ﬁeld
(e.g., Gibson, 1978; Hagan & Bumiller, 1983; Kleck, 1981 ; Thomas 8. Zingraff, 1981)
have concluded that numerous methodological ﬂaws (e.g., failure to consider
contextual effects, exclusion of relevant variables) may account for many of these
inconsistencies (Miethe & Moore, 1986, p. 230).

WIthin the past decade, two changes in sentencing have been the subject of
great concem—the adoption of sentencing guidelines or determinate sentencing and
enactment of laws requiring mandatory minimum terms of incarceration for certain

offenses or types of offenders. Disenchantment with rehabilitation and the concern

2

about sentencing disparities due tojudicial discretion, coupled with demands for'law
and order“ in the 19703, gave rise to a nationwide movement toward deterrninacy
in sentencing. Between 1970 and 1980, 12 states revised their penal codes along
these lines, and a number of others were considering similar changes (Cullen,
Gilbert, 8. Cullen, 1983). In 1982, Pennsylvania enacted a sentencing guidelines
system to create statewide uniformity and consistency in sentencing. The guidelines
have been somewhat successful in reducing disparate sentences (Kramer & Lubitz,
1985); however, as in other states that have implemented such a plan, they have
been severely criticized for the increasing racial disproportionality of the prison
population (Carroll 8. Cornell, 1985, pp. 475-476). Thus, it is particularly important
to test theories explaining racial bias in states that, like Pennsylvania, have instituted
sentencing guidelines.

It has been argued that the greater and increasing black representation
among prisoners may result from their differential involvement in crime and/or the
processing blacks receive in the criminal justice system (Hindelang, 1978. p. 94).
Evidence exists to support both points of view; however, in a controversial article,
Blumstein (1982) estimated that at least 80% of the racial disproportionality is due
to the different involvement of blacks (and other minorities) in serious crimes.
Dehais (1983), on the other hand, argued convincingly that Blumstein’s estimate
rests on the assumption that the probability of incarceration following arrest does not
vary by race. He cited substantial evidence to show that this premise is unwarranted

(e.g., Hepburn, 1978; Petersilia, 1983) and that Blumstein underestimated the extent

3

of the disproportionality to be explained. The disproportionate number of minorities
in the criminal justice system and the inconsistent research ﬁndings on racial
differences have led some researchers to take a different approach in studying the
magnitude and direction of differential treatment (Myers & Talarico, 1986, p. 367).
To address this problem, more recent studies (Nardulli, 1979; Thomson & Zingraff,
1981) on sentencing have sought to locate different sources of disparity and
discrimination in sentencing. In short, research on sentencing has begun to
demonstrate a growing awareness of broader structural and contextual variables that
' determine the kind oftreatment offenders receive. ltis presumed thatthe magnitude
and direction of unequal treatment can neither be predicted nor understood without
identifying the social circumstances under which it is likely to occur (Myers &
Talarico, 1986, pp. 367-368).

Unnever and Hembroff (1988) noted that although most studies have
modeled the judges’ sentencing decisions by taking case and defendant attributes
into account,1 few of these attempts have been guided by any structural- or
contextual-based theory of decision-making processes. Thus, researchers have
contended that they have not been able to specify exactly what the conditions are
that inﬂuence the likelihood that judges will or will not discriminate against racial!

ethnic minorities. A review of the literature revealed that most studies in the area of

 

1See Bernstein, Kelly, and Doyle (1977a); Bernstein, Kick, Leung, and Schultz
(1977); Chiricos and Waldo (1975); Hagan, Hewitt, and Alwin (1979); Hall and
Simkus (1975); LaFree (1985b); Lizotte (1978); Meithe & Moore (1985); Peterson
and Hagan (1984); Swigert and Farrell (1977); Thomberry and Christensen (1984);
Tiffany, Avichai, and Peters (1975); and Welch, Gruhl, and Spohn (1984a).

4

criminal justice are dominated by a focus on individual-level case processing—in
particular, how “extra-legal” factors such as race, social class, and gender inﬂuence
court decision making (for reviews see Gottfredson 8. Gottfredson, 1988; Hagan,
1974; Hagan 8. Bumiller, 1983). The problem, however, is that the theoretical
signiﬁcance of these studies, especially at the macro-level, has not been well
developed (Sampson, 1993). In addition to the theoretical bias in favor of individual-
Ievel explanations of criminal case processing, there is little research on the
structural context of crime control in general. Liska (1987) argued that most macro-
level research in this area has focused on deterrence (i.e., the effect of crime control
on crime rates). Only recently have sociologists used collectivities as the unit of
analysis and examined how crime control patterns are inﬂuenced by social
structures (e.g., Liska, 1992; but see contextual studies by Myers & Talarico, 1987;
Sampson, 1986). There are excellent ethnographies (e.g., Emerson, 1969) and
historical case studies on macrosocial aspects of crime control (e.g., Erikson, 1966),
but these ”illustrate rather than test sociological perspectives on crime control"
(Liska, 1987, p. 68).

Another problem besides the lack of research at the macrosocial level is the
argument that criminal justice research lacks theoretical initiative (Hagan, 1989, pp.
116-117). For example, much criminal justice research in the 1970s and 1980s
derived its theoretical initiatives either from consensus or conﬂict theories of society.
These broadly framed theories have been described as being useful in stimulating

work that concerned the inﬂuence of legal (e.g., prior record and seriousness of

5

offense) and extralegal variables (e.g., race and gender) on criminal justice
outcomes. A consensus theory of social relations predicted a powerful role for legal
variables, reﬂecting the inﬂuence of broadly shared societal values in the
punishment of criminal norm violations. Conversely, a conﬂict theory of social
relations predicted a substantial role for extralegal variables, reﬂecting the inﬂuence
of power imbalances in the punishment of crimes that posed threats to existing
power relationships.

In essence, while the larger debate that organized discussions of consensus
and conﬂict theories in the social sciences prove useful in stimulating and framing
much of the early research on criminal justice operations, the results of these studies
have not offered much support for eithertheory (Hagan, 1989). Whereas consensus
theories led researchers to expect the inﬂuence of legal variables, such as offense
seriousness and prior record, to be strong and persistent, the results ofthis research
found the inﬂuence of these variables to be moderate and inconsistent. Whereas
conﬂict theory led researchers to expect the inﬂuence of extralegal variables, such
as race and class, to be substantial and pervasive, the results of this research found
the inﬂuence of these variables to be modest and uncertain. The literature reviews
in this area of study vary somewhat with regard to these summary statements (e.g.,
see Wilbanks, 1987; Zatz, 1984), but the larger point nonetheless holds-neither
consensus nor conﬂict theory has generated large-scale empirical support. More
speciﬁcally, it is concluded that consensus and conﬂict theories do not provide

sufﬁcient attention to the structural relationships that emerge from a joining of

6
organizational and political forces in the direction of criminal justice decision-making
processes.

Using a conﬂict/Weberian-based perspective, this researcher attempted to
develop and test a macro-level framework by examining the relationship among
racial composition, urbanization, and court bureaucratization, and the relationship
of these three variables to punishment. Rather than viewing each theory as
competing against each other, some contemporary researchers have advocated an
integration that employs variables from seemingly different causal models. The
rationale behind integrating theories is that any pure theoretical statements are
partial explanations and, therefore, can be strengthened and enhanced by the
integration (Elliott, Ageton, & Carter, 1979; Johnson, 1979). Because integration
incorporates the relative strength of each theory, it is also believed to increase its
explanatory power (Shoemaker, 1984). Weberian theories suggest that court size
and location inﬂuence differential treatment, whereas conﬂict theories propose that
in highly bureaucratized courts, sentences will depend on factors not explicitly
construed as legally relevant, namely, the relative power and status of the offender.
In less bureaucratized courts, a reliance on offender status or power will be relatively
less common or absent (Myers & Talarico, 1986, p. 369). A court’s context,
therefore, strongly inﬂuences the ways in which cases are selected, heard, and
disposed, while its social structure is consistently associated with differences in rates
of crimes, the degree of procedural formality, and the administration of punishment.

As a result, where offenders live affects how their cases are disposed and the

7
severity of the sentences they receive. ConSistent with this proposition, F eld (1991)
believed that mral judges exercise their discretion in a manner that results in race,
gender, and class differences (p. 162).

Of the three variables described, urbanization has been identiﬁed in the
literature as an important contextual source of unequal treatment. Weber (1947)
associated the formal rationalization of social life with urbanization and
bureaucratization and argued that abstract rules would supplant more traditional
methods of dispute resolution as laws became increasingly rational and functionally
specialized. Presumably, urban courts would be more formal and bureaucratized,
emphasize rationality and efﬁciency, and punish on the basis of legally relevant
factors such as present offense and prior record. By contrast, rural courts would be
less bureaucratized and sentence on the basis of extralegal considerations.

Although there has been a long-standing interest in urbanization and its
correlate, bureaucratization. there have only recently been empirical studies of their
actual effect on sentencing (Austin, 1981; Hagan, 1977; Kempf & Austin, 1986;
Myers & Talarico, 1986, 1987; Pope, 1976; Tepperman, 19866). Hagan found that
differential treatment of racial minorities was more pronounced in rural courts than
in bureaucratized urban ones. ,Tepperman reported that rural juvenile courts treated
female offenders more leniently than males, but that gender differences declined
with urbanization. Austin found that rural criminal courts considered social
background factors, while urban courts adhered to a more legalistic model of

sentencing. Myers and Talarico reported that urbanization and social context affect

8

criminal court sentencing, while Kempf and Austin assessed the effect of the
urbanization factor on the sentence outcome and concluded that racial disparity in
sentencing was revealed more clearly when separate analyses were conducted
within levels of urbanization. These studies support Weberian expectations that
similarly situated offenders may be treated differently based on their locale and that
differential treatment is more prevalent in rural settings and declines with
urbanization and bureaucratization. Finally, the research thus far has indicated the
need to further explore this area of study because it appears that the urban-
bureaucratic factor may be more complex that we have been led to believe.

This study was therefore premised on the proposition that as the level of
urbanization increases, the courts become more bureaucratized and less likely to
discriminate against offenders regardless of their race, gender, or class. In other
words, in bureaucratic courts, most offenders may experience the same criminal
processing because efﬁciency, not individualized justice, becomes the major goal
of the organization. No court is exempt from discriminatory or unequal justice;
however, this researcher explored the possibility that discrimination and disparity in
sentencing may diminish as bureaucratization increases. Because bureaucratization
is likely to be greater in the city, where the population is larger and heterogeneous,
minorities should experience more lenient treatment when they are sentenced in
urban rather than suburban and rural courtjurisdictions. The ﬁrst research question

addressed in this study was: What are the effects of race, urbanization, and court

9
bureaucratization on a sentence of prison versus no prison (e.g., probation or some
other sentence alternative such as ﬁnes, restitution, and so on)?

The literature revealed that there may be other important contextual
characteristics, such as the size of the subordinate population and crime rates,
which may be confounded with urbanization. It has been suggested that the
differences attributed to urbanization may actually be due to these contextual
factors, which, while linked with urbanization, are nonetheless conceptually and
empirically distinct phenomena (Myers & Talarico, 1986). The researcher, therefore,
attempted to address this methodological shortcoming by including some of the
contextual variables that have been omitted from other studies. It is known from
previous research that race and other extralegal variables (e.g., gender, age, and
so on) have been useful in explaining some of the variation in sentencing. However,
the second research question addressed in this study was: To what extent do
contextual variables (e.g., percentage black, percentage poor, percentage urban,
and so on) predict the probability of an IN sentence (e.g., jail or prison) across
counties in the State of Pennsylvania which have not been predicted by individual
factors? WIII contextual factors be more helpful in providing additional information
that has not been obtained from the examination of individual factors? Or are some
individual factors more important in some contexts than others? In any event, this
researcher examined the relationship between the type of punishment the offender
received and the location in which he or she was sentenced. It was assumed that
individual-level (race, age, gender) and contextual (percentage black, percentage

poor, percentage urban, and so on) variables are inextricably linked; therefore, they

10
should be considered simultaneously if one wishes to understand sentencing and,
by implication, other social responses to crime and punishment.

Finally, in this study, seriousness of offense and prior offenses were
considered as control variables. These legalistic variables, which are speciﬁed as
the inﬂuences on decisions in consensus theories, might explain (i.e., make
spurious) relationships between the other independent variables and the INIOUT

decision.

CHAPTER II

THEORETICAL PERSPECTIVES

Various theoretical perspectives have been used by criminologists to explain
the disproportionate number of blacks and minorities in the criminal justice system.
These theories all differ in their explanations, and they generally include consensus
or conﬂict perspectives. Borrowing from the work of organizational theorists, some
criminologists have recently turned to structural organizational perspectives to
explain the decision-making processes of the judiciary. These Weberian models
(Chambliss & Seidman, 1971; Reiss, 1974; Tepperman, 1973; Weber, 1946, 1947,
1954) are some of the approaches used by researchers to help understand how
discretion is exercised in court bureaucracies. In sum, these theoretical frameworks

form the basis of this study and are presented below.

IbeCmsensusEeLsnemile
Consensus theories of law are said to explain disparities in the sentencing
process in terms of differential criminal involvement, viewing criminal law as a set of
codiﬁed norms equally applied for all law violators. The severity of punishment
imposed for crimes varies directly with the importance of the law violated (Durkheim,

1954, 1973). This perspective presumes that sanctions are imposed by the system

11

12
of criminal justice primarily in relation to the seriousness of crimes committed, with _
the most serious and violent offenders receiving the most punitive sanctions. Finally,
it is presumed that the administration of criminal justice treats most defendants
equally, without regard for their social standing or other personal characteristics
(Bridges & Crutchﬁeld, 1988, p. 700).

According to consensus theories, racial differences in imprisonment occur
because minority males, particularly black males, violate the law more frequently and
commit more serious crimes than members of other racial groups (Blumstein, 1982;
Hindelang, 1978; Langan, 1985). The racial distribution of offenders imprisoned is
approximately equal to the racial distribution of persons arrested because no
signiﬁcant racial differences exist in treatment of the accused following arrest—that
is, at prosecution, conviction, sentencing, or in actual time served in prison. Finally,
these theories predict that blacks and other minorities will be imprisoned at
disproportionately higher rates than whites in regions where blacks have
disproportionately higher rates of arrest for serious and violent crimes than whites
(e.g., see Harries, 1980; Webster, 1978).

WOW
S I I' I E l l'

Conﬂict theories explain criminal punishment in terms of hegemony by
dominant social classes, viewing disparity in punishment in terms of economic and
political inequality within society. Racial disparity is not only found in imprisonment

rates, but there is differential treatment in how criminal cases are processed and

13

disposed in the legal system. Minorities are more frequently pretrial detainees, more
likely to plead guilty, more likely to receive longer sentences upon conviction, and
more likely to be incarcerated for crimes than whites who commit similar crimes
(Christianson, 1980a, 1980b; Lizotte, 1978; Quinney, 1970, 1974). In the most
widely shared explanation, imprisonment and other aspects of the legal process are
institutional mechanisms that have integral ties to the economic and political order
in society.

Marxist theorists argue that the use of criminal punishment is closely
associated with economic stratiﬁcation. They reason that economic elites use the
legal institutions to control and manage society’s problem population, typically the
chronically unemployed and persons living in extreme poverty. In societies and
communities characterized by rigid economic stratiﬁcation and heavy urban
concentration of poor, elites are likely to use the administration of criminal justice to
enforce laws that preserve the economic order (Chambliss & Seidman, 1971;
Humphries 8 Greenberg, 1981 ; Jacobs, 1978; Rusche & Kirchheimer, 1968; Spitzer,
1975, 1981). Criticizing this line of reasoning, Bridges and Crutchﬁeld (1988, pp.
700-701) argued that gross racial differences in imprisonment should be expected
in those regions of the country and historical periods where levels of black/white
economic inequality among the very poor and black concentrations in central cities
are most pronounced. Like the conﬂict theorists, they predicted that minorities will

be imprisoned at disproportionately higher rates than whites in those regions where

14
blacks are more heavily represented among the very poor and more heavily
concentrated in urban areas than whites.

Conﬂict theorists also have argued that the minority threat is likely to vary in
relation to the size of the minority population, with large minority populations
substantially more threatening to whites than small populations (Barth & Noel, 1972;
Blalock, 1957, 1967; Brown & Fugitt, 1972; Brown & Warner, 1992; Frisbie &
Neidert, 1976). Studies that have shown the evidence of threat have revealed a
greater fear of crime among people who migrate from cities to smaller communities
(Kennedy 8 Krahn, 1984). Other studies have revealed that some suburban and
rural residents migrate from cities not only because of the fear of crime, but for racial]
ethnic considerations, including school desegregation (Bosco & Robin, 1974;
Dobriner, 1963; Lord & Catan, 1977), and that there is usually greater precaution
taken by nonurban residents relative to their awareness of crime in their
communities (80993, 1971). These ﬁndings suggest that discrimination is more
likely to be found in communities and regions where the minority population is
largest and presents the most serious political threat. lf sentence disparity reﬂects
racial disparity, offenders who are sentenced in jurisdictions where minority
percentage is high compared to the white population will experience harsh
treatment.

In summary, conﬂict theorists have suggested that racial differences In the
sentencing process typically are produced by three aspects of the legal process that

afford white defendants less severe punishments than blacks, even among persons

15
committing similar types of crimes (Bridges & Crutchﬁeld, 1988, p. 702). First, racial

discrimination may occur overtly in legal decisions, with judges and other ofﬁcials
often granting white defendants more favorable dispositions than blacks (Davis,
1969; Ouinney, 1970). Second, class biases can enter into the legal processing of
cases based on economic resources, including contextual factors such as the
chronically unemployed and persons living in extreme poverty. Third. racial
discrimination in legal processing can also be produced by organizational or
institutional aspects of criminal justice that have the unintended effect of ensuring
that minority defendants receive less favorable dispositions than whites (Lizotte,
1978; Swigert 8. Farrell, 1977).

Organizational constraints on courts may also disadvantage minority
defendants because court personnel, given organizational limits on the number of
persons who can be processed, may unintentionally target minority defendants for
processing (Bernstein et al., 1977; Blumberg, 1967; Sudnow, 1965; Swigert 8.
Farrell, 1977). Under conditions of prison crowding, ofﬁcials may accord highest
priority to imprisoning offenders whose behavior conﬂicts most with the norms
enforced bythe agencies and who ﬁt traditional stereotypes of serious criminals—that
is, violent black offenders—while discretion in sentencing and parole processes may
also penalize minorities. In jurisdictions and regions of the country where legal
processing is individualized and punishment is discretionary, it is said that ofﬁcials
often set prison terms according to the offender's background and living

environment.

16

As an alternative to the conﬂict perspective, researchers recently have turned
to Weberian models to explain how organizational and contextual factors can be
used to explain how courts process cases through the criminal justice system. The
inﬂuence of these variables on the sentencing process is further explained in the
next section on Weberian perspectives and the bureaucratic justice model.

I] Illl . II I' IE I' I
W

According to the Weberian perspective, courts are viewed as bureaucracies.
The bureaucratic justice model argues that courts are bureaucracies and the size
and location of the court’s jurisdiction (urban, suburban, and rural) will inﬂuence the
differential treatment of offenders. The larger the court, the more likely it will become
bureaucratized. The organizational goal then becomes ”efﬁciency” and not
”individualized justice.” The literature indicates there may be some
misunderstanding about this notion of bureaucratic justice where one is led to
believe that ”justice" is sacriﬁced in the name of efﬁciency. However, in her
discussion of bureaucratic justice, Pollock (1994, pp. 138-139) argued that this is not
the case.

In the bureaucratic justice model, each individual case is seen as only one of
many for the professionals who work in the system. Because each case is part of
a workload, decision making takes on a more complicated nature. Each case is not
separately tried and judged, but is linked to others and processed as a part of a

workload. The bureaucratic system ofjustice is seen as developing procedures and

17

policies that, although not intentionally discriminatory, may contribute toa perception
of unfairness. For instance, a major element in bureaucratic justice is the
presumption of guilt, while the ideal of our justice system is a presumption of
Innocence. Judges, district attorneys, and even defense attorneys approach each
case presuming guilt, and they place a priority on achieving the most expeditious
resolution of the case. This is the basic rationale behind plea bargaining, whether
recognized or not; the defendant is assumed to be guilty, and the negotiation is to
achieve a guilty plea while bargaining for the best possible sentence—the lowest
possible is the goal of the defense, while the highest possible is the goal of the
prosecutor. Plea bargaining is consistent with the bureaucratic value system
because it is the most efﬁcient way of getting the maximum punishment for minimum
work.

Descriptions of bureaucratic justice such as the following allow for the fact
that efﬁciency can be tempered with other values and priorities:

The concept of bureaucratic justice . . . provides the most persuasive account

of how the participants in the criminal process reconcile legal and

bureaucratic forces. 'Bureaucraticjustice unites the presumption ofguilt with

the operational morality of fairness.” . . . All participants in the criminal

process behave as if a person who is arrested is probably guilty.

Nevertheless, the coercive thrust of the presumption of guilt is softened

somewhat by the operational morality of fairness that leads the participants

to make certain that defendants get neither more nor less than is coming to

them—that defendants, in other words. get their due. (Scheingold, 1984, p.

158)

Scheingold was referring to the practice of judges, defense attorneys, and

prosecutors (hereinafter called the courtroom workgrou ps) who adapt the system to

their personal styles ofjustice. Moreover, in almost all cases, there may be general

18

consensus on both sides as to what is fair punishment for any given offender.
Instead of describing the justice system as a system that practices the presumption
of innocence and takes careful steps to determine guilt, what may be more realistic
is to characterize it as a system wherein all participants assume guilt, take superﬁcial
steps to arrive at the punishment phase, and operate under a value system that
allocates punishment and mercy to offenders according to an informal operating
standard of fairness.

This study does not focus on the behaviors of judges, prosecutors, and
defense attorneys nor plea bargaining in the sentencing process. However, in courts
identiﬁed as bureaucracies, this informal operating standard of fairness involves the
use of ”normal penalties.” which may help to explain why sentencing in bureaucratic
courts is different when compared with courts that are not bureaucratized. In
seeking to produce individualized sentences, courtroom workgroups employ what
have been labeled normal penalties (Sudnow, 1965, p. 254). Normal penalties are
sentences based on the usual manner in which crimes are committed and the typical
backgrounds of the defendants who commit them. The decisions made by
courtroom workgroups develop typical sentences of what punishments are
appropriate for given crime categories. It is within these normal penalties that
individualization is said to occur and that upward and downward adjustments are
made. Typical sentences are not used mechanically; rather, they are said to guide

sentencing.

19

Next, sentencing involves a two-stage decision-making process. After
conviction, the ﬁrst decision is made. whether to incarcerate or grant probation to the
defendant (i.e., the INIOUT decision). The second stage of the decision is
determining how long the sentence should be. This process illustrates how different
courtroom workgroups employ varying concepts of normal penalties. From one
courtroom to the next, there are important differences in the threshold for granting
probation. Once it has been decided that a defendant should be imprisoned, there
are important differences in the factors used to determine the length of that
sentence. Stated another way, courtroom workgroups tend to look at the same set
of general factors in passing sentence; however, there is no uniformity in the relative
weights that are assigned these general or individual factors (Neubauer, 1988, pp.
362-364). Therefore, if discrimination does result from bureaucratic justice, this
consequence is unintended because the statuses ascribed to offenders are not used
as the criteria for sentencing. Efﬁciency, not individualized justice, becomes the
primary goal (Hagan, 1977, p. 598).

A Conﬂict theorists have criticized thebureaucratic justice model, holding
opposing views as to whether discrimination is intentional or not. Reiss (1974), for
example, addressed the consequences of court bureaucratization, arguing that the
discretion given to agents of the law opens the door to unequal treatment,
particularly when the limits of discretionary power are unclear. The position
presented above, however, suggests something otherwise. In his position, Reiss

(1977) did not speciﬁcally indicate whether the variability of bureaucratized justice

20

is differentially targeted at minority-group offenders; however, Chambliss and
Seidman (1971) did. These authors began with the assumption that "the tendency
and necessity to bureaucratize is far and away the single most important variable in
determining the actual day-to-day functioning of the legal system" (p. 468). They
argued:

The large number of persons brought before the municipal courts for minor

transgressions of the law leads to almost complete automatic sentence for

certain types of offenders. Furthermore, even for the more serious offenses

the pressure to make the decision expeditiously (which is in large part a

carry-over from the heavy burden created by the large number of minor

offenders handled) leads to the judges relying heavily on the advice of

'specialists'—in this case, probation and parole ofﬁcers who make

presentencing reports on offenders before the court for sentencing. (p. 468)
Under these circumstances, Chambliss and Seidman argued, institutional patterns
of discrimination against the poor are inevitable.

Hagan (1977, pp. 598-599) reviewed a variety of studies that offer good
reasons why sentencing disparities may or may not occur in urban court
bureaucracies. Combined with these conﬂicting arguments is a consensual view
that either outcome is undesirable. Thus, Reiss predicted a variability in
bureaucratized justice that Chambliss and Seidman agreed is discriminatory,
whereas others (i.e., Tepperman, 1973; Turk, 1976) predicted a uniformity in
sentencing that, on one hand, is seen as impersonal, and on the other, deﬂecting
attention from the disadvantaging character cf the laws themselves. The apparent
irony of this situation is that researchers agree that the consequences of

bureaucratization are bad; however, there is disagreement about what these

consequences actually are.

21

Conﬂict theorists have hypothesized that in highly bureaucratized courts.
sentences will depend on factors not explicitly construed as legally relevant-namely,
the relative power and status of the offender. In less bureaucratized courts, a
reliance on offender power and status will be relatively less common or absent
(Myers 8 Talarico, 1986, p. 369). There is a large and expanding body of literature
that has looked for systematic links between the characteristics of offenders and the
sentences they receive. Much of this work has included an effort to test the conﬂict
theory of crime. The results of these studies have been inconsistent; whereas a
variety of studies have found these relationships to be substantial (Lizotte, 1978;
Swigert & Farrell, 1977), others have found them to be contingent on particular
circumstances (Hagan, Bernstein, & Albonetti, 1980).

In response to such studies, conﬂict theorists increasingly concede the point
that relationships between extralegal characteristics and court outcomes are neither
as large nor consistent as is frequently assumed. Instead, they argue that such
ﬁndings are actually quite consistent with the perspective which suggests that a
discriminatory court system would be more harmful than good for the administration
of justice (Turk, 1976). It is argued that ”ruling classes have a general interest in
promulgating and reproducing the stability of the social order as a whole, and an
important way of achieving this is by somehow ensuring that the severity of
sanctions ought not signiﬁcantly be correlated with social class” (Beirne, 1979, p.

373).

22
II I' II I' IE I'

The literature on sentencing reveals that conﬂict models are the most fully
developed perspectives and they have rendered the most substantial empirical
support (McCarthy, 1990, pp. 325-326). These perspectives are also more useful
because they focus on structural characteristics of sanctions, particularly the extent
to which unemployment and large urban populations inﬂuence the use of social
control (Rusche 8 Krrchheimer, 1939; Turk, 1969). Although less empirical attention
has been devoted to the effect of the economic and social composition of the
population, there is also evidence that the level of unemployment, the size of the
racial minority, and economically disadvantaged populations may also inﬂuence the
use of incarceration (Brenner, 1976; Carroll 8 Doubet, 1983; Greenberg, 1977; Hale,
1988; lnveriaty 8 McCarthy, 1988; Jacobs, 1978). It is noted that these results have
not been achieved invariably; therefore, many questions remain about the extent and
nature of the effects of contextual factors and their relationship to sentencing
(McCarthy, 1990. p. 326).

The recent integration of conﬂict theories with Weberian-based perspectives
has allowed criminologists to look at contextual variables such as urbanization and
its strong correlate, court bureaucratization, and their inﬂuence 0n sentence
dispositions. The relationship between sentencing decisions and the type of court
environment was speciﬁcally considered by Bullock (1961), Eisenstein and Jacobs
(1977), Hagan (1979), Austin (1981), Kempf and Austin (1986), and Kramer and

Steffensmeier (in press). All ﬁve studies questioned the inﬂuence of urbanization on

23

racial disparity in sentencing; however, the evidence on bureaucratization and the
relative power of dominant groups is less convincing. For example, Kramer and
Steffensmeier, using 1985-1987 Pennsylvania sentencing guidelines data, looked
at a contextual factor using percentage Republican and percentage Democratic.
Their contention was that communities with a high percentage of voters registered
as Republicans signiﬁed a more conservative or ”law and order' social environment.
They also looked at other county contextual factors such as racial mix, caseload,
and case mix; however, they found that none of their county contextual variables
was noteworthy in explaining dispositional decisions.

Kempf and Austin (1986), using 1977 Pennsylvania guidelines data,
assessed the effect of the urbanization factor on the sentence outcome. They found
that racial disparity in sentencing was revealed more clearly when separate analyses
were conducted within levels of urbanization. Kramer and Steffensmeier (in press)
argued that while it is possible that the guidelines instituted in Pennsylvania in 1982
may have eliminated the effect of the urbanization factor, the importance of the latter
in affecting sentence outcomes as reported in the Kempf and Austin study may have
been due to their failure to include adequate statistical controls for offense
seriousness and prior record. The research thus far has indicated the need to
further explore this area of study because it appears that the urban-bureaucratic
factor may be more complex than we have been led to believe.

Another positive characteristic of conﬂicttheories isthat they give researchers

a realistic perspective on sentencing in a social context which appears to be abstract

24

in consensus theories. Earlier works on unequal treatment lend strong support for
the involvement theory, which argues that minority groups dominate the criminal
justice system because they commit the majority of the crimes (Blumstein, 1982;
Hindelang, 1978; Langan, 1985). Conﬂict and stratiﬁcation theories, on the other
hand, suggest that disproportionate numbers of minorities in the criminal justice
system exist for reasons beyond the personal lives (e.g., lack of education,
unemployment) of offenders. These theories, however, provide a theoretical
framework that looks at the social structure where the goal of the socially dominant
group is to remain in a relative position of power over subordinate groups. As a
result, minorities are arrested, prosecuted, and incarcerated in greater numbers,
especially in urban areas where the minority composition is greater.

A question might be raised as to whether this study integrates theoretical
perspectives or is testing competing theories. Although conﬂict theories are based
on different assumptions about society than Weberian theories, there may be some
broader conﬂict in the social stnrcture between the classes that the integration of
both perspectives can explain. For example, there may be bureaucratic constraints
(from the judicial system) that mediate or interact with larger social forces to
counterbalance or reduce the inequities of a system that favors those in control.
Weberian perspectives argue that as the level of court bureaucratization increases,
discrimination and disparity should decrease in place of other organizational goals
that take precedence (i.e., efﬁciency, or in other instances, the power and status of

the defendant). However, in their study, Myers and Talarico (1986, p. 387) found

25

that urbanization emerged in its own right, independently of bureaucratization, as a
contextual determinant of differential treatment. In illustrating the importance of one
contextual determinant of differential treatment, they encouraged the identiﬁcation
of other contexts (for example, economic inequality, crime rates) that could play
similar roles.

Apart from replication, Meyers and Talarico (1986) recommended that the
next task for researchers is to discover the intervening mechanisms, such as
contextual variables which translate population characteristics and level of
urbanization into decisions that, in the aggregate, reveal distinct patterns of
differential treatment. It is therefore important that punishment be placed in a
broader social context without abandoning those factors that are at the individual
level of analysis (such as race). It is believed that these variables—both individual
and contextual—are inextricably linked, and they must be considered simultaneously
if one wishes to understand sentencing and, by implication, other social control
responses to crime.

Conﬂict and Weberian perspectives are therefore two important theories
whose integration should be further examined because they can provide useful
information about unequal treatment in the criminal justice system. Kempf and
Austin (1986) argued that bureaucratization is likely to be greater in the city and,
more important, the size of the urban minority population should give it greater
relative power than its nonurban counterpart. This proposition could be used to

explain some of the inconsistent ﬁndings on race and urbanization. Because most

26

urban areas have a higher minority composition, one might expect this group to
receive differential treatment for a number of reasons. First, minority groups are
usually the dominant social group in large urban areas, often referred to as
”chocolate cities,"1 where they have some political leadership. This leadership may
be reﬂected in the manner in which defendants are treated in the criminal justice
system, especially where there is a substantial racially mixed police force and court
personnel. A racially mixed criminal justice system may reduce or, in some
instances, eliminate the racial disparities that might otherwise exist. These
contextual factors may explain why black defendants sometimes receive more
lenient treatment than their white counterparts when they are sentenced in urban
versus suburban and rural areas.

A ﬁnal contradiction worth noting is that disproportionately longer prison
sentences are sometimes imposed on white offenders in urban counties. Several
factors are believed to explain this result (Myers 8 Talarico, 1987, p. 384). Lawsuits
alleging discrimination in prison conditions and in sentencing decisions, coupled with
sympathy toward the disadvantaged, could foster reverse discrimination. One also
cannot rule out the operation of regionally based patemalistic attitudes. For
example, it was found that southern judges expected less of blacks and were more

tolerant of their criminality. By virtue of membership in a caste that has traditionally

 

1A ”chocolate city” is any citywhere minority groups are greater in p0pulation and
in most instances are in political control of the local government. Detroit, Michigan;
Washington, DC; Atlanta, Georgia; and East St. Louis, Illinois, are cities
representing examples of this concept.

27

considered itself racially and economically superior, judges may hold whites to a
higher standard of behavior. Their transgressions may be seen as particularly
reprehensible and as warranting relatively more severe punishment (Bernstein et al.,
1977). In large cities, social characteristics may not be as important as seriousness
of the offense and the defendant’s prior record. The organizational goal of most
urban courts is to expeditiously dispose cases, and, as Weberian-based perspective
suggests, it is in this process that they become more bureaucratized. In major urban
areas, the socially dominant group are minorities; this again may therefore explain
why differential treatment based on race is not as prevalent as it is in suburban and
rural areas.

The integration of conﬂict-Weberian theories in this study is important
because it will be used to examine the process in which individual and contextual
factors predict the outcome of a sentence, based on the racial composition of the
jurisdiction in which the offender is sentenced, whether the court is bureaucratized,
and whether thejurisdiction is located in an urban, a suburban, or a rural area in the

State of Pennsylvania.

IIII'IISI'D”

Borrowing from the ideas of Miethe and Moore (1986), this researcher argues
that racial bias in criminal processing may be the artifact of the model selected to
detect racial differences. Speciﬁcally, although commonly overlooked in past
research, it is argued that the selection of a particular analytical model also implies

some fundamental assumptions about how race, urbanism, and court

28

bureaucratization should inﬂuence sentencing decisions in the State of
Pennsylvania. Using a statewide sample of convicted felons, additive and interactive
terms are estimated to determine which variables (e.g., extralegal, legal, and
contextual) are the best predictors of sentence outcomes. This study supports the
conclusions drawn from similar studies2 which argue that ”additive” or ”main
effects,“ commonly used in past research, suppress the nature and magnitude of
racial and/orjurisdictional differences. To address thisvspeciﬁcation error, interactive
terms for extralegal (e.g., race, gender, age) variables, and the variable representing
the level of urbanization (e.g., court size), were estimated. The ﬁndings from these
results are discussed later with the implications for this and future studies on

disparities in sentencing.

Mess:

The theoretical integration of conﬂict-Weberian theories described above will
allow one to examine the context in which sentencing varies across counties in the
State of Pennsylvania. The major thesis of this study was that as courts become
more bureaucratized they are less likelyto discriminate against offenders regardless
of their race or other extralegal factors (e.g., gender and age). In other words, In
bureaucratic courts, offenders may experience the same criminal processing as
offenders in courts that are not characterized as bureaucratic, because efﬁciency,

not individualized justice, becomes the major goal of the organization. Because

 

2See reviews by Myers 8 Talarico (1987), Myers (1986), and Miethe and Moore
(1986)

29

bureaucratization is more likely to be greater in the city, where the size of the
subordinate population is larger, blacks sentenced in urban jurisdictions should
experience more lenient treatment than when they are judicially processed in
suburban and rural court jurisdictions.

The ﬁrst research question addressed in this study is: What are the effects
of racial composition, urbanization, and court bureaucratization on a sentence of
prison (IN) versus no prison (OUT)? The context in which one is sentenced may be
inﬂuenced not only by one’s race, but also by the location in which the offender is
sentenced.

In this study, a model was developed that looks at the effects of how race and
other individual-level factors (e.g., gender, age, prior record, and seriousness of the
offense) predict whether an offender is sentenced to prison (IN) versus no prison
(OUT). It has been learned from previous research that race and other extralegal
(gender and age) and legal (severity, prior record [history], and number of current
convictions) factors have been useful in explaining some of the variation in
sentencing. However, the second research question raised here is: To what extent
do contextual variables (e.g., percentage black, percentage poor, percentage
unemployed, and so on) predict the likelihood of a prison versus no—prison sentence
which has not been explained by these individual-level factors?

The purpose here was to study the ”hidden” and unexplored arena of
macrostruclural variations in sentencing, especially those decisions that result in a

sentence of incarceration versus probation. To test these ideas, this researcher

30

attempted to develop a theoretical framework that links micro- and macro-level
organizational and political forces to help understand the important kinds of
variations that occur in sentencing across these social contexts. The researcher
assumed that contextual and individual-level (e.g., race, gender, age, severity, prior
record, number of current convictions) variables are inextricably linked; therefore,
they should be considered simultaneously if one wishes to understand sentencing
and, by implication, other social responses to crime and punishment. The reward
for meeting this challenge is a clearer explication of the relationships of race and the
other predictor variables to sentencing (Myers 8 Talarico, 1987, p. 387).

In establishing this point, the extralegal (race, gender, and age), legal (prior
record, seriousness of the offense, type of offense, and current convictions), and
contextual (size of the subordinate population, inequality, and crime rates) variables

are described below as they relate to the conﬂict-Weberian theories of punishment.

Emmallaﬁahlﬁs

Race

The inability to document the claim of discriminatory treatment has been
illustrated in previous empirical studies of racial differences in criminal sentencing.
Earlier researchers generally concluded that black defendants are victims of
disparate treatment. However, although there are numerous exceptions, the bulk
of recent evidence now suggests that race and other extralegal factors have little
direct effect on sentencing practices once controls are introduced for seriousness

of the charge, prior criminal record, and other legally relevant factors (see Hagan,

31

1974; Hagan 8 Bumiller, 1983; Hagan et al., 1979; Kleck, 1981; Lizotte, 1978;
Spohn et al., 1982; Unnever et al., 1980). Several authors (Bernstein et al., 1970;
Carter 8 Clelland, 1979; Unnever et al., 1980) have noted the claim that racial and
class discrimination in criminal sentencing is an essential proposition in various
theories on the administration of justice (e.g., labeling theory, conﬂict theory, and
radical criminology). However, empirical research frequently has failed to
demonstrate such biases in sentencing decisions. For instance, blacks have been
found to be treated with greater harshness in some situations, but with greater
Ienience in others (Famworth 8 Horan, 1980; Kleck, 1981; Peterson 8 Hagan,
1 984). Other researchers have concluded in their reviews that there is little evidence
of racial bias, or, ifdifferential treatment does exist, it is transmitted indirectly through
various presentence and processing outcomes (Hagan, 1975; Hagan 8 Bumiller,
1983; Lizotte, 1978; Unnever et al., 1980).

Based on the conﬂict-Weberian theory of punishment, this study supports the
claim that the offender’s race inﬂuences sentencing due to stereotypical images
relating race to the location in a social group that is thought to account for a
disproportionate amount of crime (Black, 1976; Burke 8 Turk, 1975; Chambliss 8
Liell, 1966; Green, 1961, 1964; Skolnick, 1966; Stinchcombe, 1963; Sutherland,
1949; erlick, Gehlick, 8 Watts. 1975). Miethe and Moore (1986, p. 231) challenged
this “caste-like” speciﬁcation of race that underlies stratiﬁcation theories, arguing that

in comparison to other blacks and white counterparts. those blacks who are given

32
the most severe sanctions are single, live in urban areas, have a prior felony record,
and commit multiple and serious offenses.

According to the prevailing stereotypes cited above, it is black offenders who
are most likely to be viewed as ”dangerous” by criminal justice ofﬁcials. This
researcher therefore hypothesized that black offenders, compared to white
offenders, will be incarcerated more often and receive more severe sentences when
they have a prior record and are sentenced in suburban and rural areas, where they
most often are perceived as a threat. Second, race is also believed to have
interactive effects with each of the predictor variables used in this study. The
challenge, as previously stated, is to predict the forms they will take. Where
appropriate, the hypotheses for the relationships between race and these variables

are described below.

GendeLansLAge

The literature on sentencing shows that, until recently, an adequate
explanation for gender differences in the sentencing process eluded scholars
because they tried to explain this phenomenon using conﬂict or labeling theories.
Because each theory was developed to explain class or race differences for men,
neither sufficed to explain gender differences (Daly, 1989a, p. 137). With regard to
race, gender, and sentencing, the race and sentencing literature largely tells about
racial differences in sentencing men becausewomen are usually no more than 10%

to 15% of court defendants.

33

Gruhl, Welch, and Spohn (1984) analyzed gender differences and the
likelihood of receiving an incarceration sentence for black, white, and Hispanic
(presumably Mexican-American) defendants. Adjusting for relevant controls, they
found that the average incarceration was lowest for Hispanic women (35%), followed
by black women (45%) and white women (47%). Incarceration rates for men were
higher and more similar for both black and white men (56%) and Hispanic men
(57%). With regard to gender, urbanization. and court bureaucratization, previous
research has suggested that there are inter— and intrajurisdictional differences in
sentencing that stem from court organizational factors, a state’s sentencing policies,
and judicial backgrounds (Brosi, 1979; Eisenstein 8 Jacobs, 1977; Gaylin, 1974;
Hogarth, 1971; Levin, 1977; Ryan, Ashman, Sales, 8 Shane-Dubow, 1980; US.
Department of Justice, 1984, 1985, 1987a, 1987b). But how such variability might
affect gender-based sentencing disparities is not well known (Daly, 1979, pp. 14-15).
Pope’s (1975) comparison of outcomes in urban and rural courts revealed more
favorable sentences for women in urban courts, but it showed no gender differences
in rural courts. Gruhl et al. (1981) found that male and female judges exhibited
similar conviction and sentencing patterns with one exception-male judges were
less likely than female judges to sentence female defendants to prison. Daly
(1989b, p. 15) found some interjurisdictional differences in her study, but there was
as much judicial variability within each jurisdiction. Both courts served jurisdictions

that were neither urban nor rural, but somewhere in between.

34

Trends characterizing unequal treatment by age (Austin, 1981; Pope, 1976)
and gender are less consistent. For example, Tepperman (1973) discovered that
gender differences declined as juvenile courts became more urbanized. Pope, on
the other hand, found that gender differences for adults increased with urbanization.
Finally, Austin's study supported the earlier work of Pope (1976) and Hagan (1977)
and found that suburban, compared to urban, courts sentenced older offenders to
prison in disproportionate numbers, notwithstanding the absence of correlated legal
variables, whereas rural courts sentenced both nonwhite and older offenders to
prison in disproportionate numbers notwithstanding the absence of such variables.
It was concluded that urban courts adhered to a more legalistic model of sentencing
than suburban and especially rural courts.

This researcher examined the relationship among race, gender, and age.
Based on the theoretical and empirical implications of conﬂict theory, black
offenders, regardless of their gender or age, are more likely to receive an IN
sentence compared to their white counterparts. It was predicted that younger white
females, compared to black females, are less likely to be incarcerated, whereas
older black male offenders are more likely to, receive an IN sentence. WIth regard
to urbanization and court bureaucracy, as the level of urbanization increases, the
race, gender, and age of the offender should diminish, because organizational

factors, such as efﬁciency, become more important than individualized factors.

35
Whigs
ELIQLBede 1

Conﬂict theorists would argue that black offenders are likely to have a history
of criminal behavior because they are more likely to be pretrial detainees, more likely
to plead guilty, more likely to receive longer sentences upon conviction, and more
likely to be incarcerated than whites who commit similar crimes (Christianson,
19803, 1980b; Lizotte, 1978; Quinney, 1970, 1974). One of the major criticisms of
previous research in this area is that much of the research on racial differences in
sentencing has failed to control adequately for legally relevant variables, most
important, for variables of prior record and seriousness of the offense (see Hagan,
1974; Kleck, 1981). Most of these studies used crude and imprecise measures of
prior record and offense severity, or they failed to control simultaneously for both
variables.

Hagan (1975) and Kleck (1981) were especially concerned about controls for
prior record. Kleck claimed that racial differences obtained in capital punishment
studies outside the South were due to their failure to control for prior criminal record.
He supported this contention, however, by mistakenly stating that Hagan’s
reanalysis of Nagel’s (1969) study showed that ”racial effects shrank” (Kleck, 1981,
p. 786) when prior record was controlled. Instead, Hagan reported an interaction
between race and prior record. The difference between racial groups increased and
retained statistical signiﬁcance for the group with at least one prior conviction,

whereas it declined and lost signiﬁcance for the group with no prior conviction.

36

Hagan did suggest that stricter controls over the number of previous convictions
might eliminate the racial bias found in noncapital cases, but such speculation is a
shaky foundation for a conclusion about the effect of controls when empirical
evidence shows an interaction effect (Kempf 8 Austin, 1986, pp. 32-33).
Consistent with earlier research, prior record is a strong predictor of sentence
outcome, particularly ifthe offender is black. Because blacks are more likely to be
arrested and convicted of crimes more often than whites, this should increase the
probability that they will receive harsher sentences. It was therefore hypothesized
that, during the sentencing process, a prior record of felony convictions will increase
the probability of an IN sentence because it was assumed that having a prior record
suggests the offender is more likely to be involved in future criminal behavior. It is

also relied upon to determine sentence severity (Albonetti, 1991).

S 'l [C 'lIDti IS 'I]

The severity of the sanction is a legally relevant variable and, as with prior
record, is also considered a strong predictor in determining dispositional decisions.
As such, the seriousness of the offense is expected to exert an inﬂuence on the
sentence outcome. Because it is assumed that blacks are convicted of more serious
and violent types of offenses, this group is more likely to be subjected to differential
treatment in comparison to their white counterparts committing similar offenses.
This is especially true when they are sentenced in suburban and rural areas, where

they are perceived as a social threat to those in control. It was therefore predicted

37
that the more serious the offense, the greater the likelihood the offender will receive

an IN sentence.

II I [D ID 'I' [C '1']

During the presentencing stage, the offender may have other multiple charges
such as possessing a weapon or aggravating circumstances during the commission
of a felony, such as in the case of an armed robbery. The offender may have pled
guilty and/or been found to be guilty of these'additional charges by a jury or bench
trial at the time of sentencing, but the sentencing has not been completed. In this
study, it was hypothesized that the greater the number of current convictions at the
time of sentencing, the more likely the offender will receive a harsher sentence.
Multiple convictions also indicate the severity of the crime because, ifthe crime was
not as serious, additional charges may have been plea-bargained away during the

initial or preliminary hearing and therefore not considered at the time of sentencing.

Contextuallaﬂahles
After 50 years of research on whether there are racial or ethnic disparities in
sentencing, there is only one generalizable ﬁnding: sometimes judges discriminate
and sometimes they do not (Unnever 8 Hembroff, 1988, p. 53). Some authors have
suggested that this inconsistent ﬁnding is attributable more to the methodological

ﬂaws and omissions of the research than to the absence of uniform discrimination

38

on the part ofjudges.3 Although there has been no shortage of attempts to model
judges’ sentencing decisions by taking case and defendant attributes into account,‘
few of these attempts have been guided by any contextual-based theory or decision-
making processes. Thus, researchers have not been able to specify exactly what
the conditions are that generate the likelihood thatjudges will or will not discriminate
against minority groups. In short, research on sentencing has begun to demonstrate
a growing awareness of broader contextual factors that determine the kind of
treatment defendants receive. Several researchers have identiﬁed urbanization as
an important source of unequal treatment.5 In some of these studies (Hagan, 1977;
Myers 8 Talarico, 1986; Tepperman, 1.973), researchers identiﬁed court
bureaucratization as a strong correlate of urbanization and distinguished it from
urbanization to estimate its unique effect On sentencing.

Although noteworthy for considering the broader context of sentencing, these
studies have not deﬁnitely speciﬁed the relationship between urbanization and

punishment. Most notably, it is argued that previous researchers have ignored

 

3For example, see Hagan and Bumiller (1983); Kleck (1981); Klepper, Nagin, and
Tierney (1983); Pruitt and WIlson (1983); Spohn, Gruhl, and Welch (1981), and
Unnever (1982) for reviews of these methodological concerns.

4See, for example, Bernstein, Kelly, and Doyle (1977); Bernstein, Kick, Leung,
and Schultz (1977); Chiricos and Waldo (1975); Hagan (1975); Hagan et al. (1979);
Hall and Simkus (1975); LaFree (1985); Lizotte (1978); Miethe and Moore (1985);
Peterson and Hagan (1984); Swigert and Farrell (1977); Tiffany et al. (1975);
Thomberry and Christensen (1984); and Welch et al. (1984).

5See Austin (1981), Hagan (1977), Kempf and Austin (1986), Laub (1983), Meithe
and Moore (1986), Myers and Talarico (1986, 1987) and Pope (1976).

39

important county characteristics (e.g., size of subordinate population, income
inequality, percentage minority, percentage unemployed, and crime rates) that may
be confounded with urbanization. Thus, the tendency to attribute observed
differences to urbanization per se could be misleading. In actuality, it is suggested
that these differences could be due to county factors that, although linked with
urbanization, are nonetheless conceptually and empirically distinct phenomena
(Myers & Talarico, 1986, p. 369).

Urbanization and court bureaucratization are therefore two important
variables in their study. Their causal relationship to sentencing, based on the

conﬂict-Weberian perspective, is described below.

III 'I' ISI I III'II

Urbanization has generally been regarded as one of the most important
correlates of criminality (Kornhauser, 1978). Virtually all criminological theory, to
some degree, assumes that urbanism is crucially important. In American sociology,
Wirth (1938) argued that impersonal ties, primary group structure, and the normative
consensus within urban populations are weakened considerably due to the size,
density, and heterogeneity of these populations. The progenitors of modem theory,
Thrasher (1927) and Shaw and McKay (1942), developed the idea that social
disorganization leads to a breakdown of social-control mechanisms within the urban
community, leading, in turn, to gang formation and extensive crime and delinquency.

These early notions heavily inﬂuenced later theorists (e.g., Cloward 8 Ohlin, 1960;

40
Cohen, 1955), who also focused their attention on gang delinquency in the city
(Laub,1983.p.183)

However, there exists in the literature a direct challenge to this position (see
Gordon, 1975, 1976). This challenge, referred to as ”compositional theory,”
questions the importance of urbanization and contends that differences in crime
rates across the urban-rural dimension can be attributed to differences in the
composition of the populations residing in these areas.° In other words, urbanism
itself has no major inﬂuence relative to that of other variables in accounting for
variation in urban-rural crime rates. Research has shown that crime rates are higher
in urban than in rural areas (see Harries, 1980; Webster, 1978). In addition, it is
known that blacks and other minorities tend to locate in large cities (US. Bureau of
the Census, 1972). The question arises as to whether there could be a confounding
effect between urbanism and race in accounting for variation in urban-rural crime
rates. It has been suggested that the social characteristics of the population
aggregates be examined (Quinney, 1964, p. 1). Urbanization may also be expected
to condition the effect of structural factors on social control. Quinney (1966) argued
that law enforcement is more rigid in nonmetropolitan areas as a result of greater
conﬂict between subordinates and dominants in these communities. WIth regard to
sentencing, Hagan (1977), among others (Austin, 1981; Miethe 8 Moore, 1986;
Pope, 1976), found that differential treatment of racial minorities was more

pronounced in rural courts.

 

6For a summary of these theories, see Fischer (1975, 1976).

41

A court’s context, therefore, strongly inﬂuences the ways in which cases are
selected, heard, and disposed, whereas its social structure is consistently associated
with differences in rates of crimes, the degree of procedural formality, and the
administration of justice. For instance, urban counties are more heterogeneous,
more diverse, and less stable than rural counties, where cases are disposed in a
process described as being more formal and due-process oriented. Second, urban
courts also sweep a broader, more inclusive net and encompass proportionally more
minorities and younger youths than do suburban or rural courts. Finally, the social
structure and procedural formality are also associated with more severe sanctions.
Some researchers have argued that formal urban courts are believed to sentence
similarly charged offenders more severely than do suburban or rural courts. As a
result, where offenders live affects how their cases are disposed and the severity of
the sentences they receive. Rural judges’ exercises of discretion can also result in
race, gender, and class differences (Feld, 1991, pp. 157-162).

In most studies, urbanization has been operationalized using social structural
variables, which in this study included the size of subordinate groups (e.g.,
percentage unemployed, percentage black, economic inequality, and percentage
poor) and crime rates. The causal relationship between urbanization and these

structural variables and dispositional decisions is described below.

5' [SI I' IE II'
In the literature, two distinct but parallel views have been advanced regarding

the relative size of racial minority and economically disadvantaged populations and

42

the resulting efforts at social control by elites (McCarthy, 1990, p. 328). (Both
perspectives anticipate that such efforts will increase as the relative size of the
subordinate population grows. The ﬁrst view posits that the dominant elite will
respond to the threat posed by a large subordinate population by increasing the use
of social control. The second position argues that subordinates are less effective in
resisting social control; thus, their powerlessness increases its effective use (Liska
8 Chamlin, 1984).

Studies of subordinate populations have examined the racial and economic
composition of the population using measures such as percentage minority,
percentage unemployed, percentage of families below the poverty line (percentage
poor), percentage urban, and economic inequality, as measured by the Gini Index
(Bailey, 1981; Jacobs, 1978, 1979; Jencks, 1992; Land, 190; Sampson 8 Wilson,
1993; Wilson, 1987, 1991). At present, it has been noted (Myers 8 Talarico, 1987,
pp. 25-26) that much of the research on subordinate-dominant relations has focused
on measures of social control other than conﬁnement, but the results generally have
supported the conﬂict perspective, which found that greater inequality and larger
black populations tend to foster a stronger police force (Huff 8 Stahura, 1980;
Jacobs, 1979; Liska, Lawrence, 8 Benson, 1981), greater use of deadly force
(Jacobs 8 Britt, 1979), larger police expenditures (Jackson 8 Carroll, 1981; Lizotte,
Mercy, 8 Monkkonen, 1982), higher arrest rates (Liska 8 Chamlin, 1984; Williams
8 Drake, 1980), and higher imprisonment rates (Jacobs, 1978; Joubert et al., 1981;

but cf. Bailey, 1981; Carroll 8 Doubet, 1983).

43
It was therefore hypothesized that as the size of the subordinate population
increases, so does the “perceived threat' of minorities (especially blacks), the
unemployed, and the poor. As a result, in areas where the racial composition of
blacks is high, one can expect to see more intensiﬁed social control—more
criminalization, more formal processing bythe criminaljustice system, and increased
incarceration—compared with groups that are perceived as less threatening to the

status quo (Sampson, 1993, p. 288).

Unemployment

The relationship between unemployment, crime, and imprisonment has long
interested social scientists. The common view of how unemployment inﬂuences
imprisonment emphasizes the negative consequences of economic deprivation.
Unemployment creates a stressful state that renders individuals susceptible to
criminal behavior in order to overcome their economic problems (Bonger, 1916;
Brenner, 1976). In this view, there is a direct relationship between increased
motivation to commit a crime produced by unemployment and resultant criminal
behavior. In turn, heightened rates of imprisonment during economic downturns
may stem from either a direct reﬂection of elevated crime rates (Brenner, 1976) or
a mechanism whereby social-control authorities remove members of the underclass
from the labor market to relieve economic crises (Rusche 8 Kirchheimer, 1939;
Wallace, 1981).

The few studies on how unemployment affects rates of imprisonment have

been similarly inconclusive. Although some researchers reported that rates of

44

imprisonment were invariant overtime (Blumstein 8 Cohen, 1973; Blumstein et al.,
1977), others found that imprisonment rates ﬂuctuated inversely with rates of
unemployment (Greenberg, 1977; Wallace, 1981), and still others discovered that
a direct relationship existed after controlling for changes in the crime rate (Galster
8 Scatura, 1985). However, numerous problems plagued most of these studies,
preventing any deﬁnitive conclusions because most researchers have not heeded
Radzinowicz’s (1939) early cautions that the relationship between economic
conditions and crime will probably differ across different social groups, types of
crimes, and the state of the administration ofjustice in question. In addition, most
studies of unemployment, crime, and imprisonment have been cross-sectional, an
approach that is inherently unsuited for the study of the interrelationships between
other variables because of the absence of lagged effects in such models. There
may be considerable delay between a period of unemployment and resulting criminal
activity (Parker 8 Horowitz, 1986, pp. 752-753).

Conversely, ﬁndings regarding the effect of unemployment on the level of
incarceration have been relatively consistent (McCarthy, 1990, p. 327). These
studies have varied in methodological sophistication, but it has been found that
research in a variety of settings, employing cross-sectional and longitudinal
analyses, has shown that increases in prison population do tend to be associated
with unemployment (Box 8 Hale, 1982; Brenner, 1976; Dobbins 8 Bass, 1958;
Grabosky, 1979; Greenberg, 1977; Hale, 1988; Jankovic, 1977; but cf. Parker 8

Horvvitz, 1986; Robinson, Smith, 8 Wolf, 1974; Stern, 1940; Vogel, 1975; Yeager,

45

1979). There has been speculation that sentencing practices may account for the
effect of unemployment because unemployed offenders tend to receive harsher
sentences, but there has been no Indication whether this is a result of a desire to
conﬁne dangerous offenders who lack the stabilizing communityties of employment,
to punish persons who fail to conform sufﬁciently to maintain a job, or to remove
surplus labor from the labor pool. It has been suggested that no researcher has
attempted to relate court sentencing practices to incarceration rates, so there is no
evidence that prison sentences are more frequently used in. the speciﬁcjurisdictions
suffering from high unemployment rates (McCarthy, 1989, p. 243).

The debate about the effect of unemployment has not been settled, however,
because both methodological and conceptual reﬁnements continue to yield new
interpretations of the link between unemployment and incarceration. Recent
interstate longitudinal research by Galster and Scaturo (1985) used a. variety of
controls, including law enforcement and correctional expenditures, and
disaggregated the incarceration rate to take into account variations in new prison
commitments, in returning parolees, and in conditional and unconditional releases.
The authors found that the conﬂict view was supported only in southern states,
which are considered to be in an earlier stage of economic development than other
parts of the country. On the other hand, lnverarity and McCarthy (1988) employed
national data in a 36-year longitudinal analysis of labor markets that varied in level

of competitiveness, and concluded that the Rusche-Kirchheimer formulation which

.46

proposes that the modes of labor exploitation and dominant/subordinate relations in
a population inﬂuence the use of social control.

According to conﬂict theory, the level of unemployment will have a positive
effect on incarceration as the dominant group attempts to control these who are peer
and powerless. It was therefore hypothesized that offenders who are sentenced in
jurisdictions with high unemployment rates are more likely to receive an IN sentence

than those in jurisdictions with lower unemployment rates.

W

The conﬂict theory proposes that struCtural characteristics of the population
will have direct effects on the use of incarceration apart from variation in the rate of
crime. Although a positive relationship between crime and incarceration rates might
be regarded as nothing more than a passive response to an increased volume of
activity in the criminal justice system, empirical support forthis association has been
described as somewhat weak (McCarthy, 1990, pp. 321, 326). Longitudinal and
cross-sectional studies of the inﬂuence of crime rates on levels of incarceration have
yielded inconsistent results. Some researchers have concluded that no relationship
exists between crime rates and incarceration (Blumstein 8 Cohen, 1973; Blumstein,
Cohen, 8 Nagel, 1977). Other studies have produced positive correlations (Carroll
8 Doubet, 1983; Joubert, 1981; McGuire 8 Sheehan, 1983), whereas still other
research has yielded negative correlations (Bailey, 1981; Biles, 1979; Rutherford,

1977)

47

The relationship between crime rates and inwrceration is important; however,
there is evidence that methodological ﬂaws may explain at least some of the
inconsistencies in research ﬁndings. Whereas most researchers have concluded
that the level of crime has no effect on rates of incarceration, high and increasing
rates of crime have resulted in sentencing reforms (i.e., determinate and mandatory
sentencing policies) (Carroll 8 Cornell, 1985, p. 478). IT here is a general consensus
that urbanization inﬂuences crime rates (Bureau of Justice Statistics, 1986; Krohn,
1978; Shelly, 1981), butthe implications for social control are understood less clearly
(Myers 8 Talarico, 1987). Urbanization can be expected to increase the number of
people subject to sanctions because it inﬂuences the crime rates and fosters greater
reliance on formal means of social control (Berk, Rauma, Messinger, 8 Cooley,
1981; Liska 8 Chamlin, 1984). )Vleasures of urbanization thus are used frequently
as control variables in conﬂict analysis.

In her recent study on crime rates and jail/prison conﬁnement, McCarthy
(1990, pp. 331 -332) found that all structural variables were positively correlated with
both jail and prison conﬁnement; not unexpectedly, crime rates were also
signiﬁcantly correlated with incarceration rates. Rates ofjail and prison conﬁnement
were also highly correlated; this relationship showed that counties that made greater
use of one form of conﬁnement sanction tended to make similarly great use of the
alternative form. Finally, results of the regression analysis showed that structural

factors had a signiﬁcant direct effect on the use of conﬁnement apart from the effect

of crime rates, but the effects were not uniform and varied with the type of

48
conﬁnement under consideration. Because of the continuing expectation that crime
does inﬂuence the use of incarceration in some way, crime rates are also used
routinely as control variables in efforts to assess the effect of other factors on
conﬁnement patterns (Myers 8 Talarico, 1987).

Sampson (1985), in his study on urban homicide, argued that the higher
individual-level prevalence of offending among blacks than whites has important
consequences for aggregate analysis. Quite simply, a positive relationship between
percentage black and the aggregate crime rates across areas was expected, due
only to the effects of differing composition. This result was also expected because
high offending rates for blacks as compared to whites tended to induce a positive
correlation between percentage black and crime rates; that is, the larger the black
composition, the higher the aggregate crime can be expected. The use of the
aggregate crime rates thus has important implications for an ecological inquiry into
the effects of race on crime. The most important issue is that, in the presence of
large between-group differences, the aggregate rate obscures contextual and
compositional effects. Komhauser (1978, p. 104) raised a number of research
questions in this area, such as: How do we know that area differences in crime (and
delinquency) rates result from the aggregate characteristics of communities rather
than the characteristics of individ uals selectively aggregated into communities? How
do we even know that there are any differences at all once their differing composition

is taken into account? In the case of race and crime, ecological inquiry suggests

49
that one solution is analysis of the effects of city racial composition on the criminal
behavior of blacks and whites.

This contextual approach is consistent with the basic goal of contemporary
urban economy as reviewed by Berry and Kasarda (1977, p. 13), who argued that
a fundamental assumption of the ecological approach is that social systems exist as
entities suigenen‘s and exhibit structural properties that can be examined apart from
the personal characteristics of their individual members. As such, one would
contend that there is little doubt that racial composition represents a structural or
macrosocial property of residential environment. The experience of growing up
black in a predominantly white community is certainly different from that of growing
up in a predominantly black community. An entire class of sociological theory is built
around the effects of social structure on inter-group relations (see Blalock, 1967;
Blau, 1977).

Consequently, an exact test of aggregate-level effects of racial composition
and other structural factors requires that established correlates of criminal behavior
be taken into account. As Hindelang’s (1978, 1981) research showed. there are
extremely strong differentials in offending, not only by race, but by gender and age
as well. As Hindelang (1981, p. 472) noted, the variability in criminality explained by
these characteristics is so great that it is incumbent on sociological researchers to
take them into account. This does not imply that one must construct or call upon

extant social theory to explain demographic correlates such as age (see Hirschi 8

50

Gottfredson, 1983), but rather that one not attribute to social structure what may
simply be manifestations of individual-level differences.

In short, Sampson (1985) suggested that only by controlling for individual-
level characteristics such as race, by which aggregate units differ in their
composition, will one be in a position to isolate contextual effects reasonably.
Unfortunately, few studies in criminology have been conducted within a contextual
framework. This has been largely because of the difﬁculties in collecting data on
both individual and aggregate characteristics across a sample of areas that vary on
key variables.

Although the literature on crime rates and sentencing is inconclusive, this
writer proposed that the level of crime will have a positive effect on incarceration
(e.g., IN sentence) rates. In other words, offenders who are sentenced in
jurisdictions with high levels of crime are more likely to be sentenced IN than those

who are not.

C l B l' l'

Court bureaucratization has been identiﬁed as a strong correlate of the
urbanization factor. It is only recently that researchers have begun to examine the
relationship between these two variables, and the ﬁndings from these studies have
documented signiﬁcant urban-rural differences in sentencing. Hagan (1977), among
others (Austin, 1981; Meithe 8 Moore, 1986; Pope, 1976), found that racial
discrimination was more pronounced in rural courts, whereas Kempf and Austin

(1986) found that it was more pronounced in urban courts. They concluded that

51

racial disparity is revealed more clearly when separate analyses are conducted
within levels of urbanization. Finally, there is some evidence (Austin, 1981) that
social background factors are more important in rural courts, whereas legally
relevant factors, such as prior record, play more prominent roles in urban courts.

Little has been written to demonstrate that the size or scale of an enterprise
is itself sufﬁcient to explain the emergence of a bureaucratic organization, even
where the conformity of a subordinate population is not problematic.
Bureaucratization of decision making in a large organization (in this case, courts)
emerges in Opposition to norms of “individualized” treatment, as a response to the
scale of operation and the potentialfor conﬂict among decision makers. Vlﬁth such
bureaucratization, large courts become more like one another in their decisions; their
responsiveness to ascribed differences among clients, especially differences in race
and gender, decreases; and techniques of expediting the ﬂow of clients are
perfected, without noticeable beneﬁt to the recipients of organizational services.

Wrthin a legal system that is normatively antagonistic to the standardization
of decision making, one ﬁnds an association between court size and the degree of
court bureaucratization that may be interpreted as cause and effect. For instance,
large courts, which are typically found in urban areas, are somewhat larger than
medium-sized courts, which are usually found in suburban areas; likewise, medium-
sized courts are larger than small courts, which are characteristic of rural areas.
There is some reason to expect, then, that the scale and nature Of court activities will

be qualitatively different in the largest courts, and not merely quantitatively different

52

along a monotonic size continuum from small (I.e., rural) to large (urban). Stated
differently, one may expect that once a court passes a certain level of size or activity,
it may become more unlike smaller courts than one would have predicted from a
comparison of small and medium-sized courts alone (T epperrnan, 1973, pp. 346-
351).

The theory of court bureaucratization proposes that a primary characteristic
of bureaucratized decision making is "standardization.” If urban court systems are
bureaucratized, decision making should be more consistent overtime than decision
making in mral court systems. Similarly, one should expect to ﬁnd that
bureaucratized courts are more like one another than are nonbureaucratized courts.
Because the courts ideally are expected to treat offenders as unique individuals
without respect to race, gender, or class, itwould be a difﬁcult, if not impossible, task
to “match” Offenders in different courts to assess whether the court treatment of
similar Offenders is identical or divergent.7

In sum, the available research conveys the impression that ostensibly similar
offenders are treated differently, depending on whether they are sentenced in urban
or rural courts. Although there are exceptions. most ﬁndings have supported
Weberian expectations about the consequences of bureaucratization. Differential
treatment appears to characterize rural courts, and it declines with urbanization and

subsequent court bureaucratization (Myers 8 Talarico, 1986, p. 369). This

 

7Because Tepperman (1973) did not have the data to judge whether the same
kinds of cases were customarily treated in the same way in large as contrasted with
small courts, he cautioned readers on the validity of this proposition.

53

researcher hypothesized that individual differences in sentencing (e.g., such as
one’s race and gender) and the differential treatment of Offenders should disappear
in urban courts, which are most often characterized as bureaucratized courts. It is
believed that, in these courts, most Offenders experience the same criminal
processing because efficiency, not the status of the Offender, becomes the major
goal of the organization. One should expect, then, to see differential treatment in
suburban and rural areas, where not only the status of the offender is considered at
sentencing, but also the size of the black population in the court’s jurisdiction, which,
according to conﬂict theory, is perceived as threatening and must be socially

controlled.

CHAPTER III
METHODOLOGY

Data

In this study, the sentencing guidelines data for 1983, totaling 10,596 cases,
were analyzed. The purpose of Pennsylvania sentencing guidelines, which affect
any offender convicted of a felony or serious misdemeanor after July 21, 1982, was
to establish sentencing standards in which the severity of the convicted offense and
the offender’s criminal history are the major determinants of sentencing decisions
(Kramer 8 Scirica, 1986). Guidelines sentences are established for each
combination of Offense severity/criminal history in the form of a sentencing matrix.
Also under the guidelines, dispositional or durational departures from presumptive
sentences are permissible, but the judge must justify any departure from the
guidelines with written statements outlining the circumstances behind the departure
(Kramer 8 Steffensmeier, in press). Although it is a fairly rigorous and system-
crafted sentencing system, the speciﬁc structure and scope of the Pennsylvania
guidelines, nonetheless, affords ample opportunity for the intrusion of sentencing
(T onry, 1980).

The data for this study consisted of a statewide sample of convicted felons

in the State of Pennsylvania, based on a monitoring system developed by a

54

55

commission. Each sentence given for a separate criminal transaction must be
reported to the commission. There are 67 counties and 59 judicial districts in the
State of Pennsylvania. As viable political and social entities, counties vary markedly
in demographic, political, economic, and social composition. Typically (in previous
research), the signiﬁcance of the race of the offender being sentenced is assumed,
for example, apart from the place where the sentencing occurred. An analysis of
sentencing across all these counties not only permitted an assessment of contextual
factors, but the large number of cases (N = 10,596) permitted the systematic use of
statistical controls. The data provided sentencing information on a large number of
cases, detailed information on prior record and offense severity, and information on
a number of other variables that might affect sentencing outcomes. The data base
has been described as unique because it includes the richest information in the

country for analyzing sentence decisions (Kramer 8 Steffensmeier, in press).

MaﬂablesLamLMeasnLement

Denendentlaﬂable

The dependent variable in this study was the INIOUT decision. Sentencing
is thought of as a tvvo-stage process involving two decisions. The ﬁrst decision is
whether to imprison (IN decision, which also includes jail) or not to imprison (OUT
decision, which includes probation or some other sentence alternative, such as ﬁnes,
restitution, or community service). The second decision concerns sentence length;
however, the INIOUT decision was the focus of this study. The dependent variable

was a dichotomous variable and was measured using 1 = IN (prison/jail) and 0 =

56

OUT (probation or some other sentence alternative, such as ﬁnes, restitution, or

community service). (See Table 1 for a description of variables used in this study.)

Table 1: Description of variables, coding, and summary measures.

 

J

 

 

 

 

 

 

 

Independent Variables Codes

Race Binary: Coded 1 = black, 0 = white

Gender Binary: Coded 1 = male, 0 = female ll

Age Binary: Coded 1 = older (> 25 years), 0 =
younger (s 25 years)

Prior record Criminal history score: 7-category ordinal scale
with a range of 0 to 6

Severity Severity of convicted offense: 10-category
ordinal scale with a range of 0 to 9

Convictions Number Of current convictions at time Of
sentencing, with range of 0 to 9

Percent black % of county pepulation that is black

 

Percent unemployed

 

% of county population that is unemployed with
civilian labor force '

 

Percent poor

% of county population living below the poverty
line

 

Percent urban

% Of county population living in urban areas

 

 

 

 

 

 

Crime rates No. of crimes known to police/100,000 persons
Court size Continuous variable: Coded O = small, 1 =
medium, 2 = large
Court jurisdiction Continuous variable: Coded 0 = rural, 1 =
suburban, 2 = urban
Dependent Variable Codes
INIOUT decision Binary: Coded 1 = IN, 0 = OUT

 

===J===LE====E==L

 

57
Independentlan'ahles

The independent variables in this study included extralegal variables or the
defendant’s social attributes (e.g., race, gender, and age), legal variables (severity,
prior record, number of current convictions [convictions]), and county contextual
variables (e.g., percentage black, percentage poor, percentage unemployed,
percentage urban, and crime rates). Most of the extralegal variables were coded as
dummy variables. Two dummy variables were used to measure the defendant’s
race: (1 = black, 0 = white). Gender was also dichotomized and coded as 1 = male
and 0 = female. Age was a dichotomous variable and was coded as 0 = younger
(s 25 years) and 1 = older (> 25 years).

The legal variables included prior record, measured using a weighted seven-
category ordinal scale developed by the Pennsylvania Sentencing Commission.
These criminal-history scores measure the number and severity of the defendant’s
past convictions. All felonies, as well as misdemeanors punishable by at least one
year, are included. Misdemeanors (punishable up to ﬁve years in Pennsylvania)
may total no more than two points on the criminal-history score, whereas felonies
add one, two, or three points each, depending on their severity. Severity of the
convicted offense was measured using a 10—point scale developed by the
Pennsylvania Sentencing Commission. The scale of offense severity ranged from
1 (minor theft) to 10 (murder in the third degree). The ranking of the
misdemeanors/felonies on this scale is consistent with the rankings of offenses on

most other scales of seriousness (e.g., burglary-1 vs. burglary-2). In essence, the

58

10-point severity scale ranks each statutory offense on the scale and, for certain
offenses such as burglary, subdivides the statutory classiﬁcation into multiple ranks,
depending on the speciﬁc circumstances of the crime. (Kramer 8 Scirica, 1986).
Finally, the number of convictions (convictions) was another interval-level variable,
ranging from one to nine convictions, and it equaled the number of convictions the
offender had at the time of sentencing.

The two county contextual variables in this study were urbanization and court
bureaucratization. Urbanization was operationalized using ﬁve stmctural variables:
percentage black, percentage poor. percentage unemployed, percentage urban, and
crime rates. The information for these variables was obtained from the 1983 City
W. To disentangle the effects of urbanization from those of court
bureaucratization, two aspects of court bureaucratization were considered. They
included court size and the location of the court (e.g., court jurisdiction) where the
offender was sentenced. Initially, court size was computed using the number of
probation Ofﬁcers/county, number of caseloadslcounty, and number of presentence
reports/county. This variable was computed by ﬁrst taking an average of the number
of probation ofﬁcers, caseloads, and presentence investigation reports per county.
Because the means for each of these variables were disproportionate, these scores

were standardized before they were combined to create the court size variable.1

 

1It was suggested that court size be measured using the number of probation
ofﬁcers rather than judges. The latter was found to be an inaccurate measure
because it was redundant with the urbanization factor (Hagan, 1973; Myer 8
Talarico, 1986; Tepperman, 1973). This study revealed that court size may also be
confounded with the urbanization factor (e.g., percentage black, percentage poor,

59

Court size was a continuous variable and was coded as 2 = large urban courts, 1 =
medium suburban courts, and 0 = small rural courts. Courtjurisdiction is a measure
based on the 1980 Census Bureau computation of the percentage of the county
population classiﬁed as urban, suburban, and rural.2 In this study, this variable was
also continuous and was coded as 2 = urban jurisdictions (including Philadelphia,
Allegheny, and Pittsburgh counties), 1 = suburban jurisdictions including Bucks,
Montgomery, Delaware, and Westmoreland counties), and 0 = rural counties
(including all other remaining counties).

Crime rates was another county contextual variable, measured Using rates
(per 100,000) of violent (e.g., murder, forcible rape, robbery, and aggravated

assault) and nonviolent (i.e., property-burglary, automobile theft) crimes. The

source for these data was the 1983 WWW.

Data:Analxsis_P_mceduLes
The analysis for this study proceeded from bivariate correlations to logistic

regression equations. The bivariate correlations were examined to assess the

 

and percentage urban), and that using probation Ofﬁcers as a measure of court size
may not be any different from using number of judges.

2This variable was based on the Census Bureau computation of the percentage
of counties’ population classiﬁed as urban, suburban, and rural. The values were
arrived at in the following way: counties with 33% or less of their population
classiﬁed as urban by the Census Bureau were labeled as rural; counties with a
population ranging between 34% and 67% classiﬁed as urban were labeled as
suburban; and the remaining counties. those with 68% or more of their population
classiﬁed as urban were designated as urban (Austin, 1981).

60

strength of the associations between the independent and the dependent variables.3
Because the dependent variable representingthe INIOUT decision was binary coded
(1 = IN, 0 = OUT), logistic regression was used as the multivariate procedure.

Using the standardized parameter estimate, the analysis ﬁrst assessed the
additive and interactive effects of extralegal (race, gender, age), legal (prior record,
severity, number of current convictions), and contextual (percentage black,
percentage poor, percentage urban, percentage unemployed, crime rates, court
size, and court jurisdictions) variables on the INIOUT decision. Because the
contextual variables were highly correlated, court size was selected to represent all
of the contextual variables and used as the predictor variable in the logistic
regression model.

Using a stepwise procedure, an additive equation consisting of all variables
(e.g., extralegal, legal, and contextual) was entered into the regression model. Next,
interactive equations that included some of the independent variables and the
product terms‘ were then entered into the model. The variables that did not meet
the signiﬁcance level for entry into the model were dropped. The procedure

continued to add all other variables, one at a time, and any variable deleted at one

 

‘The magnitudes of the correlation coefﬁcients were measured on a scale where
.70—1.00 was considered a strong relationship, .40-.69 was considered a moderate
relationship, and .10-.39 was considered a weak relationship.

‘The interactive terms for this analysis included (a) black x gender, (b) age x
gender, (c) black x age, (d) black x age x gender, (e) gender x court size, (f) age x
court size.

61

step was still eligible for reinclusion at a later step. The process terminated either
when new variables were entered or whenthe one to be entered was the one
dropped at the previous step. In this study, the signiﬁcance level for staying in the
model was set at p s .05. It was predicted that the direction and magnitude of these
variables would depend on the offender’s race, the size of the court, and its location.
In short, these three variables were believed to operate as signiﬁcant determinants
of the extent and magnitude of differential treatment at the time of sentencing.
Next, the choice of any analytical procedure depends on its ability to address
two problems the data can pose: sample-selection bias and collinearity (Myers 8
Talarico, 1986, pp. 374-375). In this study, there were some limitations of the data
that restricted the generalizability Of the results. First, it contained only convicted
felons. Individuals who had all charges dropped or were acquitted on all counts
were excluded. If the cases do become more sociologically homogeneous in later
stages of processing, this suppression of social variation is expected to limit the
explanatory power of social characteristics (Miethe 8 Moore, 1986, p. 221; Zatz 8
Hagan, 1985, pp. 104-105). Next, if there is differential attrition by race through
earlier stages of processing, sample-selection bias would limit substantive
inferences about the population of felony cases (Berk, 1983). The problem of
sample-selection bias was not addressed in this study; however, it is mentioned as

one of the shortcomings of this analysis.

62

In the next chapter, the ﬁndings from this study are presented. Chapter V
includes a summary, discussion, conclusions, and recommendations for future

studies on sentencing.

CHAPTER IV
PRESENTATION AND ANALYSIS OF FINDINGS
WW“ SII IE IS II'III'

The descriptive statistics for the State of Pennsylvania and urban, suburban,
and rural court jurisdictions from which the sample was taken are shown in Table 2.
In the State of Pennsylvania, offenders classiﬁed as urban residents comprised 21%
of the statewide sample, whereas 26% were suburban and 53% were rural (Census
Bureau, 1980). On the average, blacks made up 8% of the total population, and
whereas they were virtually nonexistent in some counties, they comprised 37% of
the population in others. Across counties. the percentage of unemployed ranged
from 3% to 11% of the population, while the average crime rate was 384 felony
offenses per 100,000 residents. In urban jurisdictions (n = 2,219), 88% of the
convicted offenders were males, whereas 12% were females. WIth regard to race,
surprisingly, blacks comprised 37% of this population, while whites made up 63% of
the total offender population. The mean score for the severity variable was 4.44,
with a standard deviation of 1.80. The average prior record score was 1.44, with a

standard deviation of 2.05, while the average number of current convictions was

63

 

 

 

 

 

 

 

 

64

 

 

 

 

 

 

 

 

 

 

 

 

 

8.9 .3... ...o 8... 88 8.. 85 88 8% .58
8.2. 8.28 2.8 8.8 8.8 8.4.8 8.8. 8.8 82328.2 were
8. .8 2.8 8.8 8.8 2.8 2.8 8.8 8.8 :85 e.
2.8 2.8 3.. 2.8 8.8 8.2 8... 8.8 58 4.
8.. 8.8 83 8.8 8.0 8... 8.. 8.8 88.8585 .x.
8.8 8.4 8.8 8.8 8.2 8.8 8.8. 8.2 .8055 as
8.0 . 8.. 8.0 8.. 85 8.. Rd 8.. 88228
8.. 8.. 8.. 8.. 8.8 3.. 8.. 8.. 289.28
8.. 8.8 8.. 8.8 8.. 3.... 8.. 8.8 8:88
8.8 8.8 8.8 8.8 8.8 2.8 8.8 8.8 8<
I 8... .. 8.8. I 8.8. I 8... 2858 .x.
I 8.8 .. 8.8 I 2.8 .. 8.8 2...: .x.
I 8.2 .. 8.8 .. 8.8 .. 8.8 923 .8
.. 8.8 .. 8.8 I 2.8 I 8.8 3.8.8 as
an coo—2 dd coo—2 dd cows. aw emu—2
. «88>
68.8 n s .98 88 u s 8938 I 3.8.8 u s 89: 88.2 n a 83.35 =

 

 

 

 

 

 

 

 

 

82020583 .58 .82 van .5833 .235 new w_cm>_>meeoc .O 28m on. .0. mo_.m_§m 332030 .N oBm...

65

1.36. Minorities comprised 30% ofthe urban population, while 5% were unemployed
and 16% were poor. The average crime rate in urban jurisdictions was 515 felony
offenses per 100,000 residents. Finally, the average court size was 1.43, which is
a standardized score reﬂecting the number of probation officers, the caseload size,
and the number of presentence investigation reports.

In suburban jurisdictions (n = 2,752), males also comprised 88% of the
statewide sample of convicted Offenders, while females made upthe remaining 12%.
Racially, blacks comprised 77% of the convicted offenders, while 23% were white.
In suburban areas, the average age of the offender was 27.6 years old, with ages
ranging from 17 to 75 years. The mean score for the severity variable was 3.68,
while the mean score for those with a prior record was 1.21. Most offenders in this
jurisdiction had at least one current conviction at the time of sentencing. In suburban
areas, 6% of the population were minorities, while 4% were unemployed and 6%
were classiﬁed as poor. Finally, the average crime rate in suburban areas was 421
felony crimes per 100,000 residents, and the average size for suburban courts was
.36.

The rural jurisdictions had the largest number of convicted offenders (n =
5,625). Females comprised 12% of this sample, while males made upthe remaining
88%. Eighty percent of the offenders in nIral areas were black, while 20% were
white. The average age of convicted offenders was 27.4 years. The average
offense severity score was 3.37, while the prior record score was 1.32, and the

number of current convictions was 1.27. In rural areas, 51% of the population was

66

classiﬁed as urban, and minorities comprised 4.6% of this population, while 5.3%
were unemployed and 9% were poor. The average crime rate was much lower in
rural areas, having a mean score of 315 felony offenses per 100,000 residents.
Finally, the average court size for rural jurisdictions was .14.
D 'l' Sll'l' E El I II I
andﬁnntexluaLMaﬂables

Table 3 (N = 9,943) contains the descriptive statistics1 for extralegal and
legal variables that were expected to inﬂuence the sentence outcome for the additive
and interactive equations. These variables include race, gender, age (i.e., extralegal
variables) and offense gravity score, prior record score, and number of current
convictions (i.e., legal variables). Eighty-eight percent of the statewide sample of
convicted offenders in this study were males, while the remaining 12% were
females. Wrth regard to race, of the 9,943 cases, 72% were black and 28% were
white. F ifty-three percent of the offenders were sentenced IN (state prisonfjail), while
47% of them were sentenced OUT (probation, restitution, ﬁnes, community service).
To predict the relationship between age and the INIOUT decision, age was
dichotomized (see Table 1) to distinguish older offenders (coded 1 = > 25 years)
from younger offenders (coded 0 s 25 years). WIthin the range of 0 to 1, the mean
score for age was .445. The descriptive statistics in Table 2, however, indicate the

mean age for all court jurisdictions. As indicated, the mean age for offenders

 

1This sample size is smaller than the sample size reported in Table 2 because it ‘

omits the “other“ race category (N = 653); it only includes black and white offenders
for purposes of this study.

67

sentenced in the State of Pennsylvania was 27.4, with ages ranging from 16 to 82

years. The offense gravity score was measured using a 10—point scale. This scale

ranked each statutory offense and assessed the offense type and circumstances of

the crime. The average gravity score for a given crime was 3.68, with a standard

deviation of 1.93. The offender’s prior record score was measured using a 6—point

scale, and the typical offender had at least one prior record and one current

conviction at the time he or she was sentenced.

Table 3: Descriptive statistics for extralegal, legal, and contextual variables and

the INIOUT decision (N = 9,943).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Mean SD Range ll
INIOUT .534 .499 0-1 I
Age .445 .497 0-1 I"
Gender .881 .323 0-1
Black .718 .449 0-1
Severity 3.680 1 .930 0-9 II
Prior record 1.340 2.010 0-6
Convictions 1 .320 .766 1-9
% black 8.260 11.510 0-38
% unemployed 4.880 1.080 3-11 I
% poor 9.820 4.470 4.20 I
% urban 89.290 26.890 0-100 7]
Crime rates 381.360 129.800 136-720 I
Court size .637 .930 0-2 H
Court jurisdiction .681 .800 0-2

68

The contextual variables describe the characteristics of the county or
jurisdiction where the sentencing court is located. In this study, the descriptive
statistics for the contextual variables revealed that, on the average, blacks
comprised 8% of the sample across counties, while 9.6% of the sample were
classiﬁed as poor, and 4.8% were unemployed. The average crime rate was 381
per 100,000 residents, while the mean score for the court size was .63 and the mean
score for court jurisdiction was .68.

WI || IIIIQ! II D ..

Table 4 shows the results of the bivariate analysis. showing the correlation
matrix for extralegal, legal, and contextual variables and their relationship to the
INIOUT decision. The correlations were relatively low among the extralegal (age,
gender, black) and legal (severity, prior record, convictions) variables. The largest
correlations were prior record and age ([ = .156), severity and black (n = -.145), and
prior record and gender ([ = .122). These correlations indicate that a criminal history
was positively associated with older male Offenders, while the severity of the offense
was negatively related to black offenders. Also, women were somewhat more likely
to have a prior record.

In assessing extralegal and legal variables with contextual variables, the
offender’s race and severity of the offense were the only variables that were
signiﬁcantly related to all of the contextual variables. Being black was negatively

related to percentage black in the jurisdiction, size, percentage urban, crime rate,

69

.8. u a.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000.. .80. .00.. 8.. .88. .28.- .8... .80. 20. .38. .8... .80.- 20. 80.. .80.. 008”...me
08.. .8... 8.. .20. L8.- .8.. .20. .80. L2. .8... .8...- 8.0.. .8. .80.- 8.89000
08.. .08. .8.. .28.- .88. .20. .80. .8.. .8... .08.- ..0- :0. .80.. 0358
. 80.. ..0.. .88 .08. .80. 20. ..8. .8... .88.- .08. .08.- .80. 88
08.. .30. .20. 20.- 20. .0... .88. .88. .80. .80.- .8.. .88
08.. 8.8.- .80.- 80. .80.- .08.- .08. 0.0. :0. .80. 850.08:
000.. .80. .08. .08. ..8. ..8.- 0.0. .08.- .80. .83 e.
08.. .8.- .e... .80. .80.- 0.0.- .80.- .30. 4.809.000
000.. .80. .80. .80.- 8.. .8... .08. 0.82 8.8
000.. 8.. .8..- .080. .80.. .88. 898
80.. 000..- .80.- 80. . .02. 25s
08.. .80. 80.- .8... x88
08.. .80.- L2. .808
000.. .80.. 8<
08.. 592.
swag. “an“ an: .8 a... e 88%.... 8mm 3%.”. ﬁn... .098 2...; 8... .88... 8,. 592.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

822020 h:O\z_ 2: new 8.3ng 830.908 00m .89.. ._mmo_8.xo .O. 5.8.: 02.28.00 .v 20.2-

 

70
and urban court jurisdiction. It was positively related to percentage poor in the
jurisdiction and size of the court. These correlations indicate moderate strength of
relationships.

There were positive relationships of severity to court size (n = .231),
percentage black (n = .219), and court jurisdiction ([ = .214). The correlations
between the contextual variables and the other extralegal (age, gender) and legal
(prior record. convictions) variables were either weak. virtually nonexistent, or
insigniﬁcant. Finally, in assessing the relationship between extralegal and legal
variables and the sentence outcome, they were all signiﬁcantly related to the INIOUT
decision; however, some of these correlations were relatively low. Prior record had
the strongest relationship to the lN/OUT decision (r = .370), followed by the severity
of the offense (r = .284). These other variables were also either weakly (black. 1: =
.103; gender, 1: = .107) or virtually unrelated (age, r = -.023).

BII' I'Bl CIIIII'II
Ill IIIZQIII D ..

The correlations among county contextual variables and the association of
these variables with the INIOUT decision are shown in Table 4. The correlations
were relatively high among the contextual variables (percentage black, percentage
unemployed, percentage urban, percentage poor, crime rates, court size. and court
jurisdiction); therefore, a problem of multicollinearity became apparent An
examination of these variables revealed that the strongest correlation was between

court size and percentage black ([ = .960). The next strongest correlations were

71
between percentage poor and percentage black ([ = .810), followed by court
jurisdiction and court size (L = .795) and courtjurisdiction and percentage urban ([ =
.760). Finally, crime rate was strongly correlated with percentage black ([ = .743),
percentage urban (r = .623). and courtjurisdiction ([ = .623). Ironically, percentage
unemployed was the only variable that was negatively related to crime rate (L =
-.597), whereas crime rate was moderately related to percentage poor (r = .373).

The correlations between the contextual variables and the INIOUT decision
were not as strong as their relationship with the extralegal and legal variables. The
contextual variable most strongly related with the INIOUT decision was percentage
poor (r = .122). The other variables (e.g., percentage black, percentage
unemployed, court size, percentage urban, crime rates, and court jurisdiction) had
low or virtually no relationship with the INIOUT decision.

In summary, the bivariate analysis revealed several important ﬁndings. First,
although there were weak or virtually no relationships between some of the
extralegal and legal variables, the race of the offender and severity of the offense
were the only two variables that were signiﬁcantly related to all of the contextual
variables. Extralegal and legal variables had stronger correlations with the INIOUT
decision than did contextual variables, with prior record and severity emerging asthe
variables with the strongest associations. Second, the bivariate analysis revealed
a problem of multicollinearity among the contextual variables. To address this
problem, court size was selected and used as a proxy to represent all of the

contextual variables used in this study. except percentage unemployed. The

72
bivariate analysis revealed percentage unemployed was the only variable with
moderate or weak associations with-extralegal, legal, and contextual variables.
Therefore, in the remainder of this dissertation. court size means a bureaucratic.
large urban court, in an area with high percentages of crime. black populations, and
the poor.

In the next section, using a stepwise procedure, the independent variables
and interactive terms are entered into the logistic regression model. and the
relationships between these variables and the INIOUT decision are reported. The
analysis of the logistic regression model should provide further explanations of how
the independent variables in this study inﬂuenced sentence outcomes in the State
of Pennsylvania.

BIIEIII'I'B illllllEElilf
W
Ill IIIID! II D ..

The logistic regression results for this analysis are presented in Table 5.2

Using a stepwise procedure, four main effect variables and four interactive terms

were entered into the logistic regression model. The analysis revealed that gender,

 

2The results of logistic regression analysis can fall within three categories. These
values can take on values between -1, 0, and +1, which means the standardized
parameter estimates can have a negative effect. no effect, or a positive effect on the
sentence outcome. In this study, higher or lower values are reported for both the IN
and OUT decisions. The results with negative values are associated with the
likelihood of an OUT (e.g., probation. restitution. ﬁnes. and so on) sentence,
whereas those with positive values increase the likelihood of an lN (jail or prison)
sentence.

73
prior record, and severity were the extralegal and legal variables that inﬂuenced the
sentence outcome, whereas court size and percentage unemployed were the
contextual variables that inﬂuenced the sentence outcome. The three interactive
terms that inﬂuenced the INIOUT sentence included black x gender. age x size, and
gender x size.

Table 5: Results of the logistic regression model for extralegal, legal, and con-
textual variables and interaction terms for the INIOUT decision.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Pa 2:22:33; ate 82:2)?” p-Value

Prior record ' -.569 .018 .0001
Severity -.366 .014 .0001
Gender -.205 .121 .0001
Age x size .197 .002 .0001
Gender x size .174 .063 .0001
Black x gender .162 .070 .0001
Percentage unemployed .066 .028 .0001
Court size -.053 .009 001234

0! = 1

p s .05

Gender (SEE = -.205) was the only extralegal variable that had a direct effect
on the lNIOUT sentence, and as expected, the negative standardized parameter
estimate indicated that female offenders were less likely to be incarcerated as

compared to their male counterparts. Contrary to what was predicted, the black and

74

age variables did not meet the signiﬁcance level (n s .05) to enter the logistic
regression model. As predicted, prior record and severity had a direct effect on the
INIOUT decision. In the logistic regression model, prior record (SE = -.569) had
the strongest effect. followed by severity (SEE = -.366). However, the magnitude
and direction of the standardized parameter estimates for these variables were
negative. which means that, contrary to what was predicted, offenders having a prior
record and those who have committed a serious offense (e.g., severity) may or may
not be incarcerated for the crime or crimes they commit. Convictions was another
legal variable that was expected to be a signiﬁcant predictor of an IN sentence;
however, it did not enter the logistic regression model. This ﬁnding comes as no
surprise because the bivariate correlation analysis (see Table 4) revealed that it was
virtually unrelated to all of the other variables except the severity variable ([ = .1 14).

Court size and percentage unemployed were the two contextual variables that
inﬂuenced the INIOUT decision. As previously noted, the correlations among the
contextual variables were high. Therefore, court size was selected and used as a
proxy to represent all of the contextual variables (e.g., percentage black, percentage
poor, percentage urban, crime rates, and court jurisdiction) except percentage
unemployed. The standardized parameter estimate for court size (SE = -.053) had
a negative effect on the lN/OUT decision, and as predicted, being sentenced in a
large urban court increased the probability of an OUT sentence. Conversely. the

standardized parameter estimate for percentage unemployed (SEE = .066) had a

75
positive effect on the sentence outcome. and as expected. it increased the
probability of an IN sentence.

The interactive terms that inﬂuenced the INIOUT decision included age xsize.
gender x size. and black x gender. Conversely, the magnitude and direction of the
standardized parameter estimates for age x size (SE = .197) and gender x size
(SE = .174) indicate that older males. regardless of their race. are more likely to be
incarcerated when they are sentenced in large urban courts. On the other hand, the
interactive effect between black x gender (SEE = .162) supported the expectation
that if the offender is a black male. there is a greater chance he may be
incarcerated, despite the court jurisdiction where he was sentenced.

In Chapter V the results of these ﬁndings are discussed. Also, a summary of

the study, conclusions, and recommendations for future research are presented.

CHAPTER V

SUMMARY, DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS

S l D' .

Using a conﬂict/Weberian-based perspective, this researcher examined the
relationship between race, urbanization, court bureaucracy, and punishment. The
conﬂict perspective argues that minority offenders receive longer or harsher
sentences than whites because they are considered members of a subordinate
population that is poor and powerless. The sub0rdinate population is socially
controlled by a dominant group that has the economic and political power to ensure
that laws are created to protect them from certain criminal violations, while
sentencing other groups more severely. In contrast, the Weberian perspective views
courts as bureaucracies; as the size of the court increases, it becomes more
bureaucratic, and efﬁciency becomes the most important organizational goal. In this
study, conﬂict and Weberian theories were integrated because, when separate, they
provide only a partial explanation of theoretical statements; however, when they are
integrated, their explanatory power increases.

Two research questions were addressed in the study. First, what are the
effects of race, urbanization, and court bureaucratization on a sentence of IN (prison/

jail) versus OUT (a no-prison decision, which is probation or some other sentence

76

77

alternative, such as ﬁnes, restitution, and so on)? From previous research, it is
known that extralegal (e.g., race, gender, and age) and legal (e.g., severity, prior
record, and convictions) variables have been useful in explaining some of the
variation in sentencing. However, the second research question is: To what extent
do contextual variables (e.g., bureaucratic courts and the jurisdiction where
bureaucratic courts are found in Pennsylvania) predict the likelihood of an lN
sentence (e.g., jail or prison) across counties in the State of Pennsylvania that have
not been predicted by extralegal and legal factors?

The analyses for this study proceeded from bivariate correlations to logistic
regression. The bivariate correlation matrix was then examined to assess the
strength of the associations between the independent and dependent variables.
Because the dependent variable representing the INIOUT decision was binary coded
(0 = IN, 1 = OUT), logistic regression was used in the multivariate analysis. The
Statistical Analytical System (SAS) was employed to run the logistic regression
model. Using the standardized parameter estimate, the additive and interactive
effects of extralegal, legal, and contextual variables were examined to ascertain their
signiﬁcance with regard to the INIOUT decision. A stepwise procedure was used to
enter the variables into the model, and those that did not meet the signiﬁcance level
(0 s .05) for entry into the model were dropped.

The ﬁrst research question looked at the inﬂuence of extralegal and legal
variables and their relationship to the INIOUT decision. All of the extralegal

variables appeared in the regression model; however, gender was the only variable

 

78

that had a direct effect on the INIOUT sentence. This researcher examined the
relationship between race, gender, and age, and based on the theoretical and
empirical implications of conﬂict theory, it was predicted that females would be
treated more leniently than males. On the other hand, it was expected that black
offenders, regardless of their gender or age, would be incarcerated more often
compared to white offenders. The results of this study support the conﬂict theory,
and females were treated more leniently and were less likely to be incarcerated than
male offenders. The interactive effects between black and gender further support
conﬂict expectations, and it appears that males, particularly blacks, are more likely
to be given an lN sentence than white offenders. Gender also had interactive effects
with age and court size, and as predicted, the results revealed that older males are
more likely to be sentenced lN than their younger counterparts. Thus, in the State
of Pennsylvania, if the offender is an older black male, the probability of
incarceration is greater, particularly when he is sentenced in an urban jurisdiction.

Severity and prior record were the strongest predictors of the INIOUT
decision. However, the magnitude and direction of the standardized parameter
estimates indicated that these variables increased the probability of an OUT
sentence. This ﬁnding is contrary to what was predicted because these two
variables are known in the sentencing literature to increase, not decrease, the
probability ofincarceration. There are several explanations forthisﬁnding. First, the
bivariate correlation analysis (see Table 4) revealed that, when alone, prior record

and severity were positively related to the IN decision. However, once other

79
variables are considered (e.g., race, gender, age, court size, and jurisdiction), prior
record and severity are less inﬂuential. In other words, offenders who are sentenced
in urban courts may or may not get the sentence they deserve because factors other
than prior record and severity are considered.

Next, in this study, the dependent variable, IN and OUT, was not clearly
deﬁned. An IN sentence was deﬁned as ajail or prison sentence, whereas OUT was
deﬁned as probation or some other sentence alternative such as ﬁnes, restitution,
or community service. Despite these deﬁnitions, the INIOUT decision still does not
reveal the severity or harshness of the sentence outcome. For example, an offender
sentenced in onejurisdiction may receive a six-month jail sentence (IN), whereas an
offender in another jurisdiction is given a three-year probation sentence (OUT). In
the former case, the offender was incarcerated; however, based on the length of the
sentence, one might view the jail term as less severe than the probation sentence.
In the latter case, the offender was given probation; however, one might view the
term of probation as a harsher sentence, even though he or she was not
incarcerated. In other words, in some cases an IN sentence may be a better
penalty, particularly ifthe offender accepted it as part of a sentence bargain. F lnally,
the data used in this study were collected shortly after sentencing guidelines were
implemented in the State of Pennsylvania. Therefore, it may have been too early to
assess whether they are serving the purpose they were intended to serve. The
purpose of sentencing guidelines is to reduce disparity in sentencing among similarly

situated offenders who commit similar offenses. Sentencing guidelines also are

80
intended to ensure that sentencing is based on legal factors, not extralegal factors
such as race, gender, or age. If this is true, then the Pennsylvania sentencing
guidelines are not serving this purpose, and this is reﬂected in the logistic regression
analysis. As reported, black, gender, and age did have a positive effect on the
lN/OUT decision, whereas prior record and severity did not.

The number of current convictions was the only legal variable that did not
appear in the logistic regression model. This ﬁnding comes as no surprise because
the correlations between convictions and all of the other variables, including the
INIOUT decision, were either weak or virtually nonexistent. Convictions was
described as additional or multiple criminal charges the offender incurs for which
sentencing has not been complete. Although the analysis does not reﬂect this,
multiple convictions (e.g., weapon possession, aggravating circumstances) may also
signify the severity of the crime because if the crime was not serious, these
additional criminal charges may have been plea bargained away during the initial or
preliminary hearing. It may be possible that convictions may have the same
inﬂuence on the INIOUT decision as the severity variable, and ifthis is true, this may
be why severity was retained in the regression model whereas convictions was not.

The second research question was: To what extent do contextual variables
predict the probability of an INIOUT decision that was not predicted by extralegal
and legal variables? The multicollinearity among the contextual variables made it

difﬁcult to study the effects of these variables; therefore, they did not predict the

probability of an INIOUT sentence as well as the extralegal and legal variables. To

81

address the problem of multicollinearity, court size was selected and used as a proxy
to represent large urban courts, located in jurisdictions with a high population of
minorities and the poor.

As expected, court size did inﬂuence the INIOUT decision. However,
offenders who were sentenced in urban court jurisdictions are more likely to be
sentenced to prison if they are older black males. This ﬁnding supports conﬂict
theory, but it is not clear whether it supports Weberian expectations. Conﬂict
theorists criticize the bureaucratic model holding opposing views as to whether this
discrimination is intentional or not. Reiss (1974), for example, addressed the
consequences of court bureaucratization, arguing that discretion given to agents of
the law opens the door to unequal treatment when the limits of discretionary power
are unclear. Reiss in his position, however, did not speciﬁcally indicate whether the
variability of bureaucratized justice is differentlytargeted at minority group offenders;
whereas Chambliss and Seidman (1971) did. They argued that the large numbers
of personsbrought before municipal courts for minor transgressions of the law lead
to almost complete automatic sentences for certain types of offenders. Under these
circumstances, Chambliss and Seidman argued that institutionalized patterns of
discrimination against blacks and the poor are inevitable. If this is the case for minor
offenses, then one could expect more discriminatory practices among minority
offenders for serious offenses.

Weberian theories view courts as bureaucracies, and the size and location

of the court's jurisdiction are expected to inﬂuence the differential treatment of

82

offenders. As the level of urbanization increases, courts become more bureaucratic, .
and efﬁciency, or some other form of standardized decision making, becomes the
most important goal. Therefore, offenders who are sentenced in large urban courts,
with a high concentration of minorities and the poor, should be sentenced on legal
(e.g., severity and prior record) and not extralegal (e.g., race, gender, age) factors.
The results of this study failed to support Weberian theory because it appears that,
despite the implementation of sentencing guidelines, race was a determinant for the

probability of a prison sentence.

Conclusion

The ﬁrst research question concerned the effects of extralegal and legal
variables on the lN/OUT decision. As predicted, race did inﬂuence the sentence
outcome; however, the effect was small. More important, the study ﬁndings revealed
that older offenders, particularly black males, are more likely to be incarcerated
despite the fact that in 1982 the State of Pennsylvania implemented sentencing
guidelines to eliminate racial disparity in sentencing. This ﬁnding is further
supported by the positive interactive effects of gender and age, and the negative
effects of severity and prior record. The standardized parameter estimates for these
legal variables revealed that they decreased, not increased, the probability of
incarceration.

The study ﬁndings also revealed that extralegal and legal variables were
better predictors of the INIOUT decision than were contextual variables. The

contextual variables did not inﬂuence the INIOUT decision as expected; however,

83

the high correlations among these variables made it difﬁcult to determine the extent
to which they did make a difference. These ﬁndings suggest that more research is
needed to determine whether, in fact, contextual variables are or can be reliable
predictors of the INIOUT decision. Researchers have only begun to look at
contextual variables and their relationship with individual-level variables (e.g.,
extralegal and legal). It has been their contention that individual and contextual
variables are inextricably linked. Therefore, it has been suggested that they be
considered simultaneously if one wishes to understand sentencing and, by
implication, other social responses to crime and punishment.

Because this is a recent inception, it may be too early to conclude they are
not useful in predicting sentence outcomes. For example, Myers and Talarico
(1987), in their study of Georgia sentencing patterns, found that county contextual
variables played an important role in sentencing, whereas Kramer and Steffensmeier
(in press), using the same sentencing data used in this study, found that none of
their county contextual variables were noteworthy in explaining sentence outcomes.
They attributed these ﬁndings to the possibility the guidelines that were instituted in
Pennsylvania in 1982 may have eliminated the urbanization factor, while Myers and
Talarico believed they may have been attributed to the different political and social
contexts between Georgia and Pennsylvania. Finally, these differences may reﬂect
the differences between the two analyses and the statistical controls for severity and

prior record.

34

Finally, this researcher sought to integrate conﬂict-Weberian theories to
develop and test a macro-level framework by examining the relationship among
racial composition, urbanization, and court bureaucratization, and their relationship
to punishment. Rather than viewing each theory as competing against each other,
some contemporary researchers have advocated integration, which should
incorporate the relative strength of each theory. Weberian theory argues that court
size and location inﬂuence differential treatment, whereas conﬂict theory proposes
that, in highly bureaucratized courts, sentences will depend on factors not explicitly
construed as legally relevant, namely, the race of the offender. A court’s context,
therefore, is strongly believed to inﬂuence the ways in which cases are disposed,
whereas its social structure is consistently associated with the administration of
punishment. As a result, where offenders live affects how their cases are disposed
and the severity of the sentences they receive.

The study ﬁndings supported the major tenets of conﬂict theory; however,
they rejected the propositions of the Weberian theory. In the State of Pennsylvania,
black offenders who were sentenced in urban court jurisdictions were more likely to
be given an IN sentence. It was expected that in urban courts, characterized as
bureaucratic organizations, legal, not extralegal, factors would inﬂuence the INIOUT
decision because efﬁciency, or some other form of routine decision making, would
take precedence. This was certainly not the case in this study, and the negative
standardized parameter estimates for severity and prior record only conﬁrmed the

ﬁnding that extralegal variables (e.g., race, gender, and age) had a greater effect.

85

Whether bureaucratized or not, does this mean justice is sacriﬁced for efﬁciency?
Some participants in the criminal justice process Would argue it does not. According
to Weberian theories, the concept of bureaucratic justice provides the most
persuasive account of how the judicial decision makers reconcile legal and
bureaucratic forces. Bureaucraticjustice is said to unite the presumption of guilt with
the operational morality of fairness, and the participants in this process make certain
offenders get neither more nor less than what is coming to them. In other words,
they get what they deserve (Scheingold, 1984, p. 158).

Although it appears race played a minor role in the sentencing decisions in
the State of Pennsylvania, the results of this study indicated there is still a
disproportionate number of minorities in the criminal justice system throughout the
United States. Conﬂict theorists argue whether this is true (or not) is another issue
and one worth exploring in another study. However, it appears that race inﬂuenced
the lN/OUT decision despite the inconsistent ﬁnding on urban courts, which might

suggest something different.

8 l I' I5 E l S! l'

Because the limitations of this statewide sample restrict the generalizations
that could be made to other studies, the results of this research are nonetheless
suggestive of several other substantive inferences and are brieﬂy stated. First is the
problem of multicollinearity between the contextual variables. The confounding
effects of these variables made it impossible to determine the extent to which they

inﬂuenced sentence outcomes. This problem might be resolved by disaggregating

86

the data by county and analyzing the sentence outcomes in fewer counties to
determine whether it would be an improvement of ﬁt over the model used in this
study. Should we abandon the use of contextual variables in favor of extralegal and
legal variables? Probably not. Contextual variables may not be as useful in
predicting sentence outcomes as individual-level (extralegal and legal) variables.
However, researchers must develop some method or procedures to reduce or even
eliminate the problems that are associated with this approach.

Another limitation of this study is sample-selection bias which restricts the
generalizability of the results. This bias can occur when analyzing a subsample of
the population from which some observations have been excluded in a systematic
manner (e.g., those offenders who were acquitted or their cases were not judicially
processed). Its extent varies by sample and can be corrected only by modeling all
previous selection decisions. It is recommended that future studies on sentencing
address this methodological concern by using the Heckman technique (1974, 1975).
This procedure makes it possible to correct bias partially by using this two-stage
estimation procedure (see Berk, 1983. for a discussion on how this technique is
accomplished). This procedure not only provides information about the two
decisions—type and length of sentence—but it allows us to combine this information
in a meaningful way. Unfortunately, this problem has been noted repeatedly in
criminal justice research. However, it has not been easily remedied and will persist

unless researchers address this methodological concern.

87

The recommendations listed below are philosophical and focus on some
larger problems, which are the myths that perpetuate the inconsistencies found in
research on crime and sentencing. One recommendation is that we focus on other
critical issues, one being how we deﬁne and measure crime. If we are inaccurately
measuring factors that predict criminal behavior, it does not matter how sophisticated
our analytical tools are, if the data we are gathering are biased from the start. We
can ﬁrst begin by assessing the laws that are rather arbitrary about the kinds of
phenomena that are regarded as crime and which have generally expanded and
contracted, depending on the interests of dominant groups in a social struggle for
power. How crime is deﬁned is based on politics, and the critical issue here is
whether there is a socially unacceptable and generally unknown bias in including or
excluding certain actions or inactions of others.1 There is considerable evidence

indicating that there is.2

According to the 1982 W, there were 21,012 murders
and nonnegligent manslaughters. These murders represented only a fraction of
those killed intentionally or negligently. Conservative estimates indicate that each
year at least 10,000 lives are lost to unnecessary surgeries, 20,000 to errors in

prescribing drugs, 20,000 to doctors spreading diseases in hospitals, 100,000 to

 

This discussion on the myth of crime, criminals, and crime-control policies is

taken from an article in the text on W entitled ”Crime,
Criminals, and Crime Control Policy Myths” by Robert M. Bohm, pp. 327-345.

2See Pepinsky and Jesilow (1984), Simon and Eitzen (1982), Reiman and
Headlee (1981, p. 43), Reiman (1979), Ouinney (1979, p. 62), Liberman (1974),
Mintz and Cohen (1971, pp. 25-26), and American Friends Service Committee
(1971)

 

88
industrial diseases in hospitals, 14,000 to industrial accidents, 200,000 to
environmentally caused cancer, and an unknown number from lethal industrial
products (Pepinsky & Jesilow, 1984; Reiman, 1979; Simon & Eitzen, 1982). Yet few
of these actions or inactions are deﬁned legally as murder or manslaughter. One
reason is the myth that “white-collar crime“ is nonviolent.

Because there is this illusion that white-collar crime is not violent, it is
measured differently from ”street-level crime.“ This misconception also contributes
to the skewed measurement of minorities, who are inaccurately identiﬁed as the
”dangerous class” of offenders when this, in fact, is not true. We can begin to
correct this problem by taking white-collar and corporate crime more seriously,
prosecuting offenders who commit these crimes, and including these offenses in the
Uniform Crime Reports. It has been suggested that another problem with criminal
deﬁnitions that contribute to the myths and inaccurate data collection on statistics
measuring crime is the presumption that all laws are enforced and/or enforced fairly.
Just as there is a socially unacceptable and generally unknown bias in the deﬁnition
of crime, there is a similar bias in the enforcement of the law.

One reason so few white-collar crimes are brought to light, for example, is the
inadequate enforcement mechanism. It has been suggested that regulatory
agencies, whose purpose is to prevent white-collar crime, really do not serve this
purpose. Although there is little doubt that there would be more white-collar crime
if regulatory agencies did not exist, it is not at all clear how much white-collar crime

is prevented by their existence. In any event, it is suggested that this myth can be

 

89

sustained only by ignoring the history of efforts at federal regulation of corporate
crime. Humphries and Greenberg (1981) argued that ”regulatory agencies were the
Progressive Era’s solution to the problem of controlling business in a manner that
did not legitimize capitalism by tarnishing capitalists with the stigma of criminality“
(p. 326). Similarly, Pearce (1976, p. 88) maintained that the state intentionally
created agendas responsive to the interest of big business (also see Pepinsky 8
Jesilow, 1984, pp. 66-79). Furthermore, it has been argued that, although the
prosecution of corporate crime has increased dramatically in recent years, it may
well be merely a symbolic effort to vindicate the myth that the state is neutral and to
reinforce that the law is applied uniformly to all persons (Pearce, 1976, p. 90).
Finally, Bohm (1994, p. 330) argued that the myth that crime is increasing and
that crime is an inevitable concomitant of complex, populous, and industrialized
societies is another critical issue that should be addressed. The myth that crime is
increasing also is perpetuated by the Uniform Crime Reports, which researchers rely
on to explain crime and criminal behavior. In comparing the data from the Uniform
Crime Reports with the ﬁndings ofthe Census Bureau’s National Crime Reports from
1973 to 1980, one ﬁnds a major discrepancy. The Uniform Crime Reports show a
substantial increase in the crime rate during this period, whereas the Census Bureau
statistics indicate no increase in the proportion of victims reporting the same crimes.
In some cases, the Census Bureau reported slight decreases (Paez 8 Dodge, 1982).
Therefore, a careful examination of the historical record provides no basis for the

belief that street crime, the type of crime most people fear and are led to believe is

 

90

committed by those who are minorities and poor, is rising. The truth is that people
today are in no greater danger of being robbed orbphysically hurt than they were 150
years ago (Pepinsky 8 Jesilow, 1984, p. 22; see also Ferdinand, 1977, p. 353).
The myth that crime is an urban problem was advanced in Qﬁmejnd
Wﬂnﬂhﬂmmmmﬂmmjmmm by Shelly (1931)-
who argued that when comparing crime statistics cross-culturally (cf. Sutherland 8
Cressey, 1974, p. 25), there are at least four other problems with his proposition.
First, it fails to account for the great variation in crime rates of different complex,
populous, and industrialized societies. For example, the crime rates of Japan and
West Germany are much lower than those of the United States (Martin 8 Conger,
1980; Reiman, 1979, p. 20; also see Clinard, 1978; Stack, 1984, especially Appendix
1). Second, the proposition fails to account for the great variation in crime rates
within modem, complex, populous, and urbanized nations. For example, according
to the 1984 Uniform Crime Report, the homicide rate in the United States varied from
a low of 1.0 in New Hampshire to a high of 13.1 in Texas (per 100,000 persons).
Third, the proposition fails to account for the lack of correlation between a city’s
crime rate and its population and population density. According to the 1984 Uniform
Crime Report, for example, the city with the highest homicide rate was Gary,
Indiana. The three most populated cities in the United States—New York, Los
Angeles, and Chicago—were not found among the top 10 cities with the highest

homicide rates (also see Reiman, 1979, pp. 21-23, especially Table 1). A fourth

91
problem with Shelly’s proposition is that the claim of inevitability in the social
sciences is always tenuous and suspect (Bohn, 1994, p. 330).

The myths and the problem of inaccurately deﬁning crime therefore lead to
the popular belief that some groups, particularly minorities, are not as law-abiding
as other groups (Pepinsky 8 Jesilow, 1984, p. 47). Evidence indicates, however,
that more than 90% of all Americans have committed some crime for which they
could be incarcerated (Silver, 1968; Wallerstein 8 Wyle, 1947). This observation,
however, does not deny that crime may be more concentrated in some groups, but
only that it is unlikely to be absent in others. Criminologists are aware of these facts;
however, they continue to conduct research that attributes criminality to the
involvement theory, which argues that minorities, blacks in particular, violate the law
more frequently and commit more serious crimes than other racial groups
(Blumstein, 1982; Hindeland, 1978; Langan, 1985).

Because the crimes of some people (e.g., physicians or corporate executives)
are not easily detected, or there is not as much effort exerted in detecting them as
there is for street-level crimes, these problems are believed to sustain another myth
-that most crime is committed by poor, young males between the ages of 15 and 24
(Pepinsky 8 Jesilow, 1984; Reiman, 1979). As noted, if law enforcement agencies
were able or willing to detect all crimes, the statistics that report the incidence of
crime would be more evenly distributed among the rich and poor and all age groups,
even though it may remain more highly concentrated in some groups. With regard

to age discrimination, an additional problem with the myth is that the crime rate is

92
growing much faster than either the absolute number of young people or their
percentage of the population (Reiman, 1979, p. 24).

What effect do these myths have on crime-control policies (e.g., determinate,
mandatory sentencing, and the implementation of sentencing guidelines) inthe State
of Pennsylvania and other states that have established such sentencing reforms?
These myths about crime and criminals often form the basis of crime-control policies
initiated by the now-dominant “politically conservative,“ “law and order“ ideology in
the United States. Based largely on this ideology, crime prevention and enforcement
resources recently have been expended on some of the following priorities:
mandatory sentencing, habitual-criminal statutes, increased numbers of police
ofﬁcers, more effective police ofﬁcers, changes in the Miranda warning, preventive
detention, changes in plea bargaining, changes in the exclusionary rules, and so on.
Whereas each of these policies is intended to accomplish one or more bureaucratic
goals (e.g., crime reduction, cost-eﬂectiveness, or greater efﬁciency), one is likely
to have a signiﬁcant effect on the harm and suffering experienced by the vast
majority of the American public. They do not adequately address the fundamental
social structural elements of the crime problem (cf. Bohm, 1982; Pepinsky 8 Jesilow,
1984; Walker, 1985).

A ﬁnal myth of crime-control policy to be considered is that eliminating
injustices from the criminal justice system will reduce the level of serious crime.
Eliminating injustices from the criminal justice system is certainly a worthwhile

pursuit; however, it has been argued that it is unlikely to have an appreciable effect

93
on serious crime. The causes of most crime are believed to be found in general
social arrangements and not in the operation of the criminal justice system (Bohm,
1982; Walker, 1985, p. 206).

Criminologists therefore have a responsibility to address these major
concerns before they can properly explain the reasons we have an overrepresenta-
tion of minorities in the criminal justice system. By resolving this issue, we may
resolve the problems we have in reaching consistent and conclusive results on racial
disparity in sentencing. In his discussion of crime-control policy myths, Bohm (1994)
argued that academic criminologists and criminal justice ofﬁcials are members of the
public and are partly to blame for perpetuating these myths because they should and
often do know better. He further argued that we are in the best position to dispel
these myths and inaccuracies; however, there are several reasons why we do not.
The ﬁrst reason is that many academic criminologists ﬁnd it in both their short- and
long-term interest to perpetuate these inaccuracies for interests that may be
cognitive or other members of the general public who have internalized the myths
as part of their social “reality.” To challenge the myths would be, for many, to
undermine long-established and fundamental conceptions of society. For many
academic criminologists, what has been considered here as myth simply makes
sense or attunes with preconceived ideas. To question the myths might create
cognitive dissonance.

Other academic criminologists perpetuate myths because it is in their

structural interests to do so. Platt (1975, pp. 106-107) suggested that this is

 

94

because of academic repression and cooptation. Bohm’s (1994) impression was
that prestigious university appointments and promotions in general typically go to
those academics whose work does not fundamentally challenge myths supportive
of the status quo. It appears, then, that prestigious journals rarely publish articles
that radically deviate from an accepted, often myth-laden perspective (although this
may reﬂect considerations other than ideology). Similarly, major research grants
generally seem to be available to academics whose proposals do not fundamentally
undermine privileged positions or deviate from preconceived, often myth-laden
wisdom. Whether or not myths are perpetuated because of academic repression
and cooptation. academic life is believed to generally be more pleasant for those
who do not make waves.

Finally, Bohm (1994, p. 340) argued that criminal justice ofﬁcials perpetuate
myths and inaccuracies about crime and the criminal justice system for employment
advancement, which often depends on a responsiveness to the interests of political
and economic elites. Administrators, in particular, generally are either elected to
their positions or appointed to them by political electees. Because political election
or appointment often depends on the support of political and economic elites, those
who would dispel myths that serve interests of political and economic elites are not
likely to ﬁnd support forthcoming. In the case of police ofﬁcers, the myth of
increasing crime rates, described earlier, is used to justify larger budgets for more
police ofﬁcers and higher pay (Pepinsky 8 Jesilow, 1984, pp. 16-17, 30). Third, as

was the case with the general public, myths and inaccuracies about crime and the

95
criminal justice system also provide order to the potentially chaotic role of the law
enforcement and judicial systems. They allow these criminal justice ofﬁcials to
believe that they can do the job (i.e., prevent or control crime). Finally, as was the
case for the general public, myths and inaccuracies provide criminal justice ofﬁcials
with the basis of solidarity, common purpose, and collective unity in the face of a

hostile and potentially threatening environment.

REFERENCES

REFERENCES

Albonetti, C. (1991). An integration of theories to explain judicial discretion.
SociaLEmblems. 38- 247-266.

American Friends Service Committee. (1971). Stmgglajnuusticﬁ. New York:
Hill 8 Wang.

Austin, T. (1981). The inﬂuence of court location on type of criminal sentence:
The rural—urban factor. .iQumaLQLQLiminaIJustice- 9, 305-316.

Bailey, W. C. (1981). Equality in the legal order: Some further analysis and
commentary. SogiaLELleems, 29, 51-60.

Barth, E. A., 8 Noel, D. L. (1972). Conceptual frameworks for the analysis of
race relations. SociaLEQLces, 50, 333-348.

Beime, P. (1979). Empiricism and the critique of Marxians on crime and law.
Socialﬁmblems. 26373-385.

Berk, R., 8 Ray, S. C. (1983). An introduction to sample selection bias“ ln
sociological data. WW 46- 386-398.

Berk, R., Rauma, D, Messinger, S. I... 8Cooley.T. F. (1981). Atest of the
stability of punishment hypothesis: The case of California, 1851 -1970.
Amedcanﬁmologicalﬂexiew. 46. 805-829.

Bernstein, l. N, Kelly, W. R., 8 Doyle, P. A. (1977). Social reaction to deviants:

The case of criminal defendants. AW- 42- 743-
755.

Bernstein, l. N., Kick, 5., Leung, J. T., 8Schultz. B. (1977). Charge reduction:
An intermediary stage in process labelling. SQQIaLEQLcas- 5.6. 362-384.

Berry, 3., 8Kasarda, J.'(1977). contemporacummecologx. New York:
Macmillan.

96

97
Biles, D. (1979). Crime and the use of prisons. EedotaLErooaﬂon. 9. 39-43.
Black, D. (1976). Ihobehaviomﬂaw. New York: Academic Press.

Blalock, H. M. (1957). Percent nonwhite and racial discrimination in the South.
AmoncanﬁooiologloaLBevlew- 22. 677-682.

Blalock, H. M. (1967). IooandsajbeomoiminomLouuelatlons. New York:
Wiley.

Blau, J. R. (1977). W New York: Free Press.

Blau, J. R, 8 Blau, P. M. (1982). The cost of inequality Metropolitan structure
and violent crime. AmoncanﬁoolologloaLBexlow- 41. 114- 129.

Blumberg, A. S. (1967). The practice of law as a conﬁdence game: Organiza-
tional cooptation of a profession. LamandﬁocloILBexlow, 1, 15-39.

Blumstein, A. (1982). On the racial disproportionality of the United States’ prison
Population. JoumalotCriminaLLanLandonminoloox. 13. 1259-1268

Blumstein, A., 8 Cohen, R. (1973). Atheory of the stability of punishment.
JoumaLoLCnminaliamandﬁnmmologx. 64- 198-207.

Blumstein, A. Cohen, R., 8 Nagin, D. (1977). The dynamics ofa homeostatic

punishment process. JournaLoLCdminaLLanLandonminoloov. 61, 317-
334.

Boggs, S. L. (1971). Formal and informal crime control: An exploratory study of

urban, suburban, and rural orientations. Sooiolooioalﬂuanodx. 12, 319-
327.

Bohm, R. M. (1982). Radical criminology: An explication. criminology, 19, 565-
589.

Bohm. R. M. (1994). Crime, criminals, and crime control policy myths. In M. C.

Braswell et al. (Eds.),.1ustlge._onme._and_ejhics. Ohio: Anderson
Publishing.

Bonger, W. (1916). WWW. Boston: Little, Brown.

Bosoo, J., 8 Robin, S. (1974). White ﬂight from court-ordered busing? Urban
Eduoation. 9. 87-98.

 

98

Box, 8., 8 Hale, C. (1982). Economic crisis and the rising prisoner population' ln
England and South Wales. Wipe, .11. 20—25.

Brenner, M. l-I. (1976) EstlmatlnojbosooaLcostmmonaloconomlopollcy.

110..” OII3I- -.‘I00I-l-II0I'.-Ill .00: 0|
Washington, DC: Joint Economic Committee, Congress of the United
States.

Bridges, G. S.. 8 Crutchﬁeld, R. P. (1988). Law. social standing and racial
disparities In imprisonment. Wm 66, 699-724.

Brosi, K. B. (1979).
Washington, DC: Institute for Law and Social Research.

Brown, D. L., 8 Fugitt, G. V. (1972). Percent nonwhite and racial disparity in

non-metropolitan cities in the South. Sooialﬁoionooﬁuanodx. 5.3, 573-
582.

Brown, M. C. 8Wamer, B. D. (1992). Immigrants, urban politics, and policing in
1990. AmoncanoociolooicaLBovim. oz, 293.

Bullock, H. (1961). Signiﬁcance of the racial factor in the length of prison

sentence. In R. Quinney (Ed.), WW. Boston:
Little. Brown.

Bureau of Census. (1984). WNW-.1983. Ann Arbor, MI:
US. Department of Commerce, Inter-university Consortium for Political

and Social Research.

Bureau of Justice Statistics. (1986). QnmtnaLmoIlmlzatloanJhoLImIeoﬂalos.

W. Washington. DC: U S Department
of Justice.

Burke. P., 8 Turk, A. (1975). Factors affecting postarrest dispositions: A model
for analysis. Socialﬁmblems, 22, 313-332.

Carroll, L.. 8 Cornell, C. P. (1985). Racial composition, sentencing reform, and
rates of incarceration. 1970-1980. Justicoﬂuanotlx. 4, 473-489.

Carroll, L., 8 Doubet, M. B. (1983). US. social structure and imprisonment: A
comment. criminology, 21, 449-456.

Carroll, L., 8 Mondrick, M. E. (1976). Racial bias m the decision to grant parole.
Lammoocicmaexiew- 1.1, 93-107.

 

99

Carter, R. M., 8Wilkins, L. T. (1967). Some factors in sentencing policy.
JoumaLoionmmaLLanznmmologLondﬁollceﬁoence. 56. 505-514.

Carter, T., 8 Clelland, D. (1979). A neo-rnarxist critique, formulation and test of

juvenile dispositions as a formation of social class. Socialﬁmblgms, 21,
96-108.

Chambliss, W. J., 8Seidman, R. B. (1971). W. Readings,
MA: Addison-Wesley.

Chambliss, W. J., 8 Liell, J. T. (1966). The legal process in the community

setting: A study of law enforcement. Wm 12, 310-
317.

Chiricos, T., 8 Waldo. G. (1975). Socioeconomic status an criminal sentencing:
An empirical assessment of a conﬂict proposition. Amenganﬁoclologlgal
Sexism- 4.0, 753-773.

Christianson, S. (1980a, January-February). Legal implications of racially
disproportionate incarceration rates. CriminaLLalnLﬂullotin- pp- 59-63-

Christianson, S. (1980b, November-December). Racial discrimination of prison
conﬁnement. CriminaLLaMLBulletin- pp. 616-621.

Clinard, M. B. (1978). Citieslaitbjnlocnmezlheoaseofomitzenand. England:
Cambridge University Press.

Cloward, R. 80hlin, L. (1960). WWW
dolinouontganos. New York: Free Press.

Cohen, A. (1955). Qelinoueanovs. New York: Free Press.

Cullen, F. T., Gilbert, K. E.. 8Cullen, J. B. (1983). Implementing determinate
sentencing in Illinois: Conscience and convenience. Cdminaloustioo
Bexiew. 6.1-.16

Daly, K. (1987). Structure and practice of familial-based justice in a criminal
court. Lamandﬁoolemaemew- 21, 267-290.

Davis, K. C. (1969. W. Chicago: University of Chicago Press.

100
Dehais, R. J. (1983). _Baolal_disnmoonlonalltldnonson_and_raolalolscnmlnatlon

11- 1111 ‘1‘00‘ :- ‘ 101‘ 1‘101.-.=010-. Paper
presented at the Annual Meeting of the American Society of Criminology,
Denver, CO.

Dobbins, D. A. 8 Bass, B. (1958). Effects of unemployment on white and Negro

prison admissions in Louisiana. JoumaLotﬂlminaLLaldLandﬁnmnolooy.
48- 522-525.

Dorbiner, W. (1963). W. Englewood Cliffs, NJ: Prentice-Hall.

Durkheim, E. (1954). Emmy. New York: Free Press.

Durkheim, E. (1973). Two laws of penal evolution. (T. A. Jones 8 A. Scull,
Trans.) EoonommnsiSooietv- 2. 285-308.

Eisenstein J-. &J300b H (1977) Eelonuustice:_AnomanizationaLanalvsis_ot
mminaLcouds. Boston: Little, Brown.

Elliott. D. S., Ageton, S. S, 8 Carter, R. J. (1979). An integrated theoretical
perspective on delinquent behavior. WM
Dellnouency. 16. 3-27.

Emerson, R. (1969). W. Chicago: Aldine.

Erikson, K. T. (1966). W.
New York: Wlley.

Farnworth, M.. 8 Horan, P. M. (1980). Separate justice: An analysis of race
differences in court processes. SociaLSoionooBesoann. 9. 381-399.

Farrell, R. A, 8Swigert, V. L. (1978). Legal disposition of Inter-group and intra-
group homicides. SociologloaLouanerlv. 19- 565-576.

Federal Bureau of Investigation. (1971). W. Washington,
DC: US. Government Printing Ofﬁce.

Federal Bureau of Investigation. (1981). W. Washington-
DC: US. Government Printing Ofﬁce.

Feld, B. C. (1989). The right to counsel' ln juvenile court: An empirical study of
when lawyers appear and the difference they make. JoumaloLQnminal
Lamandoriminologv. 29.

 

101

Ferdinand, T. N. (1977). The criminal patterns of Boston since 1849. In J. F.

GaIIIher&J L McCaIIIIeY(EdS)- WM
law. Homewood, IL: Dorsey.

Fischer. C. S. (1975). Towards a subcultural theory of urbanism. Amman
JoumaLoLSooiolooy. 601319-1341.

Fischer. C. S. (1976). Iboudzanoxporionoo. New York: Harcourt, Brace,
Jovanovich.

Frazier, C. E., 8 Book, E. W. (1982). Effects of court ofﬁcials on sentence
severity.

Frisbie, W. P, 8 Neidert, L. (1976). Inequality and the relative size of minority

populations. Acomparative analysis. AW- 62.
1007- 1030.

Galster, G. C., 8Scaturo, L. A. (1985). The U. S. criminal justice system:

Unemployment and the severity of punishment. JoumamLEeseaLoan
CnmoaniDolinauenov. 22. 163-189.

Gaylin, w. (1974). EaniaLjusjioe. New York: Knopf.

Gibson, J. L. (1978). Race asa determinant of criminal sentencing: A

methodological critique and case study. momma-12.
455-477. -

Gottfredson, M.. 8 Gottfredson, D. (1980). Der-MW.
Cambridge, MA: Ballinger.

Gordan, R. A. (1976). Prevalence: The race datum in delinquency
measurement and its implications for the theory of delinquency. In

M. Klein (Ed). Ibojuxenﬂoiusﬂmxstemmp 201-284) Beverly Hills
CA. Sage.

Greenberg, D. F. (1977). The dynamics of oscillatory punishment processes.
JoumalotoriminaLLamandonminolooy. 68(4). 643-651.

Gruhl, J., Welch, S. 8Spohn. C. (1984). Women as criminal defendants: Atest
for patemallsm lnLestermEoliticaLouanenx. pp. 456-467.

Hagan, J. (1974). Extra-legal attributes and criminal sentencing: An

assessment ofa sociological viewpoint LamandﬁooietLBexiow- 6. 357-
383.

102

Hagan, J. (1975). The social and legal construction of criminal justice: A study
of the pre-sentencing process. SociaLErooloms. 22- 620-637.

Hagan, J. (1977). Criminal justice in rural and urban communities: A study of
bureaucratization of justice. Sooialﬁotoes- 56. 597-612.

Hagan, J. (1979). Criminal justice in rural and urban communities: A study of
bureaucratization of justlce In S. L. Messinger 8 E. Bittner (Eds),

CriminaLLeximearbook. Beverly Hills- CA: Sage-

Hagan, J. (1989). Why is there so little criminal justice theory? Neglected
macro— and micro-level links between organization and power. JoumaLoI
WW 26(2),116-135.

Hagan, J., Bemstein, I. N., 8Albonetti, C. (1980). The differential sentencing of
white-collar offenders. AmericanﬁooiologioaLBexlew. 45, 802-820.

Hagan, J, 8 Bumiller, K. (1983). Making sense of sentencing: Areview and
critiques of sentencing research. In A. Blumstein, J. Cohen, S. E. Martin,

8M. H. Tonry (Eds) WWW
(Vol. 2, pp. 1-54). Washington, DC: National Academic Press.

Hagan, J., Hewitt, J. D.. 8 Alwin, D. F. (1979). Ceremonial justice: Crime and
punishment in a loosely coupled system. SociaLEQLces, 56, 506-527.

Hale. C. (1988, November). UnomoloymontJmon‘sonmentonthostabllitm

. . -
OI 'IIHII.A.' .II‘ I. 0| .q'I ql' I.

W. Revision of a paper presented at the Annual Meeting
of the American Society of Criminology, Montreal, Canada.

Hall, E. L., 8 Simkus, A. A (1975). Inequality in the types of sentences received
by Native Americans and whites. Qﬁmjnology, 13, 199-222.

Harries, K. D. (1980). onmoandjhoonvimnment Springﬁeld. IL: Charles C.
Thomas.

Heckman, J. J. (1974). Shadow prices, market wages. and labor supply.
Econometrica. 42. 679-694.

Heckman, J. J. (1975).Sbadmonsos.n1arket_wages_andJaboLsuppIL£-ome

.0110 - 01 1o .01‘0 -. '110' .-. '01 10 ‘ '11
Mimeographed paper. Department of Economics, University of Chicago.

103

Heckman, J. J. (1979). Sample selection bias as a speciﬁcation error.
Econometrica. 41. 153-161.

Hepburn, J. R. (1978). Race and the decision to arrest. JournaloLBesoaLann
Cnmoanooelinouency. 16. 54-73.

Hindelang, M. J. (1978a). Race and involvement“ ln common law personal

crimes. AmoncanoociolpgicaLBeﬂaw- 46. 993-1009

Hindelang, M. J. (1978b). The uniform crime reports revisited. qumaLcj
Criminal-Instinc- 2. 1 -.17

Hirschi 8 Gottfredson. (1983). Age and the explanation of crime. AmcLican
101-11086160601001.69- 552-584.

Hogarth. J. (1971). Sentencinoasaoumampmcess. Toronto: University of

Toronto Press.

Horan, P. M.. Myers, M. A.. 8 Farnworth, M. (1982). Prior record and court
processes: The role of latent theory ln criminology research. Sociological
andﬁoqaLBesearcn. 61. 40-58.

Huff, C. R. 8Stahura, J. M. (1980). Police employment and suburban crime.
Criminolmy. 11. 461-470.

Humphries, D., 8 Greenberg, D. (1981). The dialectics of crime control. In D.
Greenberg (Ed.),Qumc_and_capi1aljcm. Palo Alto, CA: Mayﬁeld.

lnverarity, J., 8 McCarthy, D. (1988). Punishment and social structure revisited:
Unemployment and imprisonment in the US, 1948-1984. Sociological

91161160!- 29- 263-279.

Jackson, P. l., 8Carroll, L. (1981). Race and the war on crime. Amcrican
Sociologicalﬁemew. 26, 290-305.

Jacobs, D. (1978). Inequality and the legal order: An ecological test of the
conﬂict model. SooiaLEroolomo. 25- 515-525.

Jacobs, 0.. 8 Britt, D. (1979). Inequality and police use of deadly force: An
empirical assessment of conﬂict hypothesis. SociaLEl’oblomo. 26- 403-
412.

Jankovic, l. (1977). Labor market and imprisonment. CnmoandﬁoolaL-lustlco.
6, 17-31.

104

Jencks, C. (1992). WWW
Cambridge, MA: Harvard University Press.

Johnson, R. E. (1977). Julenﬂeoelinouanmndjsonginunjmated
theoLetioaLaooLoaott. Cambridge, MA: Cambridge University Press.

Joubert, P. E., Picou, E. J., 8Mclntosh. W. A. (1981). US. social structure,
crime, and imprisonment. criminology, 19, 344-359.

Kempf, K. L., 8Austin, R. (1986). Older and more recent evidence on racial

discrimination in sentencing. JoumaLcLQuanmaﬂmQﬁminclcgy. 2, 29-
48.

Kennedy, L. W., 8 Krohn, H. (1984). Rural-urban origin and fear of crime: The
case for ”rural baggage.” WM, 49, 247-260.

Kleck, G. (1981). Racial discrimination in criminal sentencing. A critical
evaluation of the evidence with additional evidence on the death penalty.
AmencarLSociolooicaLBexiem. 46. 783-805.

Kleck, G. (1985). Life supports for ailing hypotheses: Modes of summarizing the
evidence for racial discrimination in sentencing. LayLanthluman
Bohaxiot- 6. 271-285.

Klepper, S., Nagin, D.. 8Tiemey, L. (1983). Discrimination in the criminal justice
system: A critical appraisal of the literature. In A. Blumstein, J. Cohen, S.

E. Martin. & M. H. Tenn! (Eds)- Bessarchomsentencingzilmearchior
Lefcrm. Washington, DC: National Academy Press.

Komhauser, R. R. (1978). ScciaLscumcsciﬁcljngucncy. Chicago University of

Chicago Press.

Kramer, J. H. 8 Lubitz, R. (1985). Pennsylvania sentencing reform: The impact

of commission-established guidelines. Qnmoondﬁollnouonoy. 31, 481-
500.

Kramer, J. H., 8 Scirica, A. J. (1986). Complex policy choices: The
Pennsylvania Commission on Sentencing. EcchaLELQbaﬁcn, 60, 15-23.

Kramer, J. H., 8Steffensmeier, D. (In press). BacodiﬂeLences-Jnsontencing;
WWW Unpublished manuscript
The Pennsylvania State University.

 

105

Krohn, M. D. (1978). A Durkheimian analysis of international crime rates. Social
EoLces- 51. 654-670.

LaFree, G. D. (1985). Official reactions to Hispanic defendants' 1n the Southwest.
JoumaLoLBesoaLdJJnonmoandoelinguencL 22, 213-237.

Land, K. C., McCall, P. L., 8Cohen, L. E. (1990). Structural covariates of
homicide rates: Are there any invariances across time and space?

AmoncamJoumaLoLSociologx-96922.

Langan, P. (1985). Racism on trial: New evidence to explain the racial
compositions of prisons and the United States. qumaLcLQdmmaLLaw
andﬂiminologx. 16. 676-683.

Laub, J. H. (1983, July). Urban, race, and crime. qumaLcLSLimcand
Dolinouonoy. pp. 183-198.

Levin. M. A. (1977). umanooliticsandjbocn'minaLcouns. Chicago: University
of Chicago Press.

Lieberman, J. K. (1974). www. Baltimore, MD:

Penguin.

Liska, A. E. (1987). A critical examination of macro perspectives on crime

control AnnuaLBeximoLSociologx. .13- 67.

Liska, A. E. (Ed.). (1992). SooiauntealondjoqaLconlrol. Albany: State
University of New York Press.

Liska, A. E., 8 Chamlin, M. B. (1984). Social structure of crime and control
among macrosocial units. AmericanJoumaLoLSociologx. 9.0. 383-395.

Liska, A. E.. Lawrence, J. J., 8 Benson, M. (1981). Perspectives on the legal

order. The capacity for social control. AmoﬂcanJoumaLoLSooiologx. 61-
413-426.

Lizotte, A. J. (1978). Extra-legal factors in Chicago’s criminal courts: Testing the
conﬂict model of criminal justice. SociaLEroblems. 25. 564-580.

Lizotte, A. J, Mercy. J., 8 Monkkonen, E. (1982). Crime and police strength in
urban setting: Chicago, 1947-1970. an. Hagan (Ed.), Quanmaﬁyc

mminaUnnoxaﬂonsjndapplications(pp.128-148). Beverly Hills, CA:
Sage.

 

106

Loftin, C., Greenberg, D. F., 8 Kessler, R. C. (1981). mccmejneoualﬂyﬁce.
WM. Paper presented at the Annual Meeting of the
American Society of Criminology.

Lord. D. J, 8 Catan, J. C. (1977). School desegregation policy and intra-social
district migration. ScciaLScicnccQuancdy, 51, 784-796.

Martin, R. G., 8 Conger, R. D. (1980). A comparison of delinquency trends:
Japan and the United States. Criminology. 16, 53-61.

Maynard. D. W. (1982). Defendant attributes in plea bargaining: Notes on the
modeling of sentencing decisions. ScciaLErcblcmc, 29, 347-360.

Maynard, D. W. (1984). The use of jail conﬁnements' m the disposition of felony
arrests..loumaLof_Q[im106L.lus1166-11,241-251.

Maynard, D. W. (1990). A micro-level analysis of social structure and social
control: Interstate use ofjail and prison conﬁnement. Jusjiceﬁuancdy, 1,
325-340.

McGuire. W., 8Shennan, R. G. (1983). Relationships between crime rates and
incarceration rates: Further analysis. JoumaloLEesearctﬂnﬁnmcand
Delinquency. 20. 73-85.

Miethe, T. D. (1984). BaoaLdrﬁeLeneeomonminaloounoeclslonsJ
OIIO- I OIO -” ‘9-3 -I' -. 3"“. II “00- 0 I"-III 00.3 I'.
Paper presented at the Annual Meeting of the American Society of
Criminology.

Miethe, T. D. (1985). Socioeconomic disparities under determinate sentencing
systems: A comparison of preguideline and postguideline practices in
Minnesota. Criminology. 23, 337-363.

Miethe, T. D., 8 Moore, C. (1986). Racial difference in criminal processing: The
consequences of model selection on conclusion about differential

treatment. SooioloaicaLQuaﬂeﬂx. 21, 217-237.

Milakovich. M. D., 8 Weis, K. (1977). Politics and measures of success in the
war on crime. InJ. F. Galliher8J. M. McCartney (Eds) Criminology;
Emmimeandorlminauaw. Homewood- IL: Dorsey Press

Mintz, M., 8Cohen, J. s. (1971). WWW
1100618161652 NewYork: Dial.

107

Myers, M. A. (1979). Offended parties and ofﬁcial reactions: Victims and
sentencing of criminal defendants. Sooiologlcalﬁmﬂeny, 20, 529-540.

Myers, M. A., 8 Talarico, S. M. (1986). Urban justice, rural injustice?
Urbanization and its effect on sentencing. Criminology, 24, 367-391.

Myers, M. A., 8 Talarico, S. M. (1987). The social contexts of racial
discrimination in sentencing. mm 3.3. 236-251.

Nagel, S. (1969). WWW Homewood,
IL: Dorsey Press.

Nagel, S., 8 Hagan, J. (1983). Gender and crime: Offense patterns and criminal

court sanctions. In M. Tonry 8 N. Morris (Eds..) 120m6_6011051i66;_A0
armoaLImdomLoﬂosoaLoh (Vol. 4). Chicago: University of Chicago Press.

Narduli, P. (1979). W. New York: Ballinger.

Neubauer, D. w. (1988). AmorioassounsandjbooriminaUustioosysiem.
Belmont, CA: Brooks-Cole.

Paez, A. L., 8 Dodge, R. W. (1982, July). Qﬁminahdoﬂmizﬂionjm
(Bureau of Justice Statistics Technical Report). Washington, DC: US.
Department of Justice.

Parker, R. N., 8 Horwitz, A. V. (1986). Unemployment, crime, and imprisonment.
Cnminology. 211., 751-774.

Pearce. F. (1976). MW. London: Pluto Press.
Pepinsky, H. E., 8Jesilow, P. (1984). My1061061£6056_60mo. MD: Seven

Locks.

Petersilia, J. (1983). WWW. Santa
Monica, CA: Rand.

Peterson, R. D., 8 Hagan, J. (1984). Changing conceptions of race and
sentencing decisions. AmoﬂoanﬁociologjoaLBeyiew, 49. 56-70.

Platt, T. (1975). Prospects for radical criminology. In I. Taylor, P. Walton, 8J.
Young (Eds), QmigaLcﬁmjnolggy. Boston: Routledge 8 Kegan Paul.

Pollock, J. M. (1994). W.
Belmont, CA: Wadsworth.

 

108

Pope, C. E. (1976). The inﬂuence of social and legal factors on sentencing
dispositions: A preliminary analysis of offender-based transaction
statistics. JournaloLQriminaLJusﬂoe. .4, 203-223.

Pruitt, C. R. (1966). Structural characteristics, population areas and crime rates

in the U. S. ioumaLoL68mInaI_Law_QnmInology_ansLEoli66_Soi6noe. 51.
45-52.

Pruitt, C. R. (1970). IhosoolaLroaIilyotonmo. Boston: Little, Brown.
Pruitt, C. R. (1974). Critiqueoﬂegaigrder. Boston: Little, Brown.

Pruitt, C. R., 8Wllson, J. O. (1983). Alongitudinal study of the effect of race on
sentencing. LayL6018ooi61yBoyi6m. .11. 613-635.

Radzinowicz, L. (1939). The inﬂuence of economic conditions on crime.

W. 33, 1-36, 139-153.

Reiman,J.H. (1979).106_£ioh_061_0606L600106.oooL961_00560:_ldoo|ooy.
Whom. New York: Wiley.

Reiman, J. H. 8Headlee, S. (1981). Marxism and criminal justice policy. Crime
an106Iinguonoy 21, 24-47

Reiss, A. J. (1974). Discretionary justice. In D. Glaser (Ed) 18606on818
oriminology. Chicago: Rand- -M.cNalIy

Robinson, W. H., Smith, P, 8Wolf, J. (1974). Illustrative projections to 1980.
Cited“ m D. F. Greenberg. (1977). The dynamics of oscillatory punishment
processes. JoumaLoIQnmlnaLLayLandﬁnmmoIogy. 56(4), 643-651.

Rusche, G., 8Kirchheimer, O. (1939). MW. New
York: Russell 8 Russell.

Rutherford, A. (1977).

W
W. Washington, DC. U. S. Government
Printing Ofﬁce.

Ryan, J. P., Ashman, A., Sales, B. D. 8Shane-Dubow, S. (1980). Amotioan
WWW New York: Free Press.

Sampson, R. J. (1985). Race and criminal violence: Ademographically

disaggregated analysis of urban homicide. 9010160036110606061 3.1.
47—82.

109

Sampson, R. J., 8Wilson, W. J. (1993). Toward atheory of race. crime, and

urban inequality. InJ. Hagan 8 R. Peterson (Eds), £20m6_601i06.00681y
Sanford, CA: Stanford University Press.

Savitz, H. V. (1978). Black cities/white suburbs: Domestic colonialism as an

interpretive idea. Annals_61_tbe.Amcrican.Academy_of_Eoli1icaLand_Social
Science. 5136. 118- 134

Scheingold, S. (1984). Ihepolmcsmﬂayrandgrder. New York: Langan.

ShaW. c. R.. & McKay. H. D. (1982). .Luyenilecclincuencyandorbanjreas.
Chicago: University of Chicago Press.

Shelly, L. I. (1981). Qtim1an1modomiza1ion. DeKalb: Southern Illinois
University Press.

Shoemaker, D. J. (1984). Incoricschclincucncy:_An_examination_ot
W. New York: Oxford University Press.

Silver. |- (1968). Introductiomnecinﬂcngectcnmeinjjrceccciem. New
York: Avon.

Simon, D. R., 8 Eitzen, D. S. (1982). Elﬁedeyianoe. Boston: Allyn 8 Bacon.

Sko'niCK. J- (1966). JusiicewiibounriaLLayLentorcememJnccmocraﬁc
M. New York: Wiley.

Spitzer, S. (1975). Towards a Marxian theory of deviance. SociaLEroblems. 22,
638—651.

Spitzer, S. (1981). Notes towards a theory of punishment and social change. In

S. Spitzer 8 R. Simon (Eds.), WWW (Vol. 2, pp.
207-229). JAJ Press.

Stack, S. (1984). Income inequality and property crime: A cross-national
analysis of relative deprivation theory. Criminology. 22. 229-257.

Stern, L. T. (1940). The effects of depression in prison commitments and

sentences. .IoumaLotonminaLLMlminologyﬂnﬂoﬂmience. 3.1.
696-71 1.

Stinchcombe, A. L. (1963). Institutions of privacy in determination of police
administration practice. AmericaoqumaLcLSociolooy.66.150-160.

110 -

Sudnow, D. (1965). Normal crimes: Sociological features of the penal codes in
a public defender's ofﬁce. SeciaLErleeme, 12, 255-276.

Sutherland, E. H. (1949). Mitocouarcrime. New York: Dryden Press.

Swigert, V. L. (1973). Effects of court size on organization and procedure.
CanadlamBeideyLotScciologyanoAntnroeolcoy. 10. 346-365.

Swigert, V. L, 8Farrell, R. (1977). Normal homicides and the law American
SociologicaLchicw. 42. 16-32.

Thomson, R. J., 8 Zingraff, M. T. (1981). Detecting sentencing disparity: Some
problems and evidence. AmericaoqumaloLSocIoIogy. 85. 869-880.

Thomberry, T. P., 8 Christenson, L. (1984). Juvenile justice decisionmaking as
a longitudinal process. SociaLEQrces, 63, 433-444.

Thrasher, F. (1927). Inegang. Chicago: University of Chicago Press.

Tiffany, L., Avichai, Y., 8 Peters, G. (1975). A statistical analysis of sentencing
in federal courts. WWI-I166. .4. 369-390.

Tonry, M. (1980). Sentencing guidelines and their effects. In A. von Hirch, K. A.

Knapp. and M. Tonry (Eds) Inecentencinocommissiomaniits
guidelines. Boston: Northeastern University Press.

Turk, A. (1976). Law, conﬂict, and order. From theorizing to theorles Canadian
BeyieyLotSocicIccyanoArnhmeelogy 13, 282-294.

Unnever, J. D., Frazier, C. E., 8 Henretta, J. C. (1980). Race difference In
criminal sentencing. SociologicaLQuanody. 21.197-205

Unnever, J. D., 8 Hembroff, L. A. (1988). The prediction of racial and ethnic
sentencing disparities: An expectation-states approach. JeurnaLof

Bcsearcb1n£nmean1ﬂelincucncy26 53-82.

U. S. Bureau ofCensus. (1972).191.0_6606115_01106_Dooul61ion;_loL1

W. Washington, DC:
U. S. Government Printing Off 06

US Department of Justice. (19843). £06606£5J0_51616_6001666[6L1051801l606
00.066610065114682. Washington, DC: Bureau of Justice Statistics.

111

U.S. Department of Justice. (1984b). 860160cinooracticeanJ3§16166 (Special
Report). Washington, DC: Bureau of Justice Statistics.

U.S. Department of Justice. (1985). EelonysentencinanJBJoijurisdictions
(Special Report). Washington, DC: Bureau of Justice Statistics.

U.S. Department ofJustice. (1987a). Sentencingoutcomeanzoielonycounsjn
1985 (NCJ 105743). Washington, DC: Bureau of Justice Statistics.

U.S. Department of Justice. (1987b). Siarejelenueuriiandjeienﬂam
(Bulletin). Washington, DC: Bureau of Justice Statistics.

Vogel, R. (1975). Unemployment and imprisonment. JeurnaLQLCriminaLLaw
and_QnminoIogy. 10197-205.

Walker, 8. (1985). W. Belmont,
CA: Brooks/Cole.

Wallace, D. (1981). The political economy of incarceration trends in late U.S.

capitalism: 1971-1977. W. 1.0, 59-67.

Wallerstein, J. S.,8Wyle, C. J. (1947). Our law-abiding Iawbreakers.
Embaiion. 25. 107-112.

Weber, M. (1946). Eseayeinmelogy. New York: Oxford University Press.

Weber. M. (1947). IbereonrcLsociaLandcconcmicoroanization. New York:

The Free Press.

Weber. M. (1954). QnJayLineconomyandcociety. New York: Simon 8
Schuster.

Webster, W. H. (1978). £20m6_i0_106.11.8._1.911. Washington, DC: U.S.
Government Printing Ofﬁce.

Welch, S., Gruhl, J., 8Spohn, C. (1984). Dismissal, conviction, and
incarceration of Hispanic defendants: A comparison with Anglos and

blacks. W 65. 257-269

Wheeler, S., Weisbard, D, 8 Bode, N. (1982). Sentencing the white-collar

offender: Rhetoric and reality. AW, 41, 641-
659.

 

112

Wilbanks W- (1987). Ibemytncielacisicriminamsticecysiem. Monterey.
CA: Brooks/Cole.

Williams, K. R., 8 Drake, S. (1980). Social structure, crime, and criminalization:
An empirical examination of the conﬂict perspective. Sociological

Quarterly. 21. 563-575.

WllIick, D., Gehlker, G., 8 Watts, A. M. (1975). Social class as a factor affecting
judicial disposition: Defendants charged with criminal homosexual acts.

Criminology. 12. 57-69.

Wirth, L. (1938). Urbanism as a way of life. AmericanJeurnamLSocielogy, 44,
1-24.

Yeager, M. G. (1979). Unemployment and imprisonment. JeumaLQLCrimmai
LamanoCriminoIogy. 2.0. 586-588.

Younger, M. S. (1979). Handboekjenﬂneauegreesien. Belmont, CA:
Wadsworth.

Zatz, M. (1984). Race, ethnicity, and determinate sentencing: A new dimension
to an old controversy. Criminology. 22, 147-171.

Zatz, M. (1985). Pleas, priors, and prison. Racial/ethnic differences' ln
sentencing. SociaLScicnceeBesearcb. 14.169-193.

Zatz, M., 8Hagan, J. (1985). Crime, time, and punishment: An exploration of

selection bias' ln sentencing research. Joumaioﬂuaniiiaﬁyeﬂiminology.
1.103-126.

 

   

nrcHIan srnr: UNIV. LIBRARIES
lllllllllllllllllllllllIllllllllllllllllllllllllll
31293015815659