 

.£... 1.

 

 

This is to certify that the

thesis entitled

DISCRIMINATION, RACE & POWER: SENTENCING
DISPARITIES IN MICHIGAN, 1998

presented by

NORMAN SAMUEL CARTER

has been accepted towards fulfillment

of the requirements for
Cmminal Justice

M . S . degree in Urban Affairs

 

 

01/2; L l 2%

C
Major professor “

Date 84302.

0.7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

MAR 18 200"

 

 

 

\‘t

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRCIDateDuapes-sz

 

—_——-v—-——-- , , ——-

 

DISCRIMINATION, RACE & POWER: SENTENCING
DISPARITIES IN MICHIGAN, 1998
By

NORMAN SAMUEL CARTER

A THESIS

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

MASTER OF SCIENCE
School ofCriminal Justice — Urban Affairs Programs

2002

ABSTRACT

DISCRIMINATION, RACE & POWER: SENTENCING
DISPARITIES IN MICHIGAN, 1998

By

NORMAN SAMUEL CARTER

A salient concern for sociologists and criminal justice policy analysts over much
of the past century has been sentencing disparity among defendants with different
attributes. After all, one only has to look at the disproportionate representation of racial
and ethnic minorities in American correctional facilities to see that something is wrong.
With the emphasis placed on the legal and extra-legal characteristics that affect
sentencing decisions, the forthcoming analyses, which are guided by conﬂict theory,
examine the sentencing dispositions carried out in 83 district courts in the State of

Michigan.

Copyright by
Norman Samuel Carter
2002

DEDICATION

Thanks mom and dad for never giving up on me. Grandma, thank you for the
prayers. I Love You. Geoffrey & Josh, study hard. D. Ormsby,
thanks for the scholarly uplift.

ACKNOWLEDGMENTS

I would like to thank my committee members for their guidance, time, and patience:
Professor Michael D. Reisi g (chair), Professor Homer C. Hawkins & Professor Charles J.
Corley.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. vii
INTRODUCTION .................................................................................... 1
RACE & SENTENCING: A THEORETICAL FRAMEWORK .............................. 2
The “Overt Racial Discrimination” Hypothesis ......................................... 3
The “Minorities Commit More Crime” Thesis .......................................... 4
The Conﬂict Perspective .................................................................... 5
REVIEW OF THE RESEARCH LITERATURE ............................................... 10
RESEARCH OBJECTIVE ........................................................................ 17
METHODS .......................................................................................... 18
Data .......................................................................................... 18
Dependent Variables ....................................................................... 19
Independent Variables ..................................................................... 20
Oﬂ‘ender Characteristics ......................................................... 20
Legal Variables .................................................................... 20
Control Variables .................................................................. 21
Statistical Procedure ........................................................................ 22
RESULTS ............................................................................................ 24
Bivariate Results ............................................................................ 24
Multivariate Results ........................................................................ 29
DISCUSSION ....................................................................................... 34
BIBLIOGRAPHY ................................................................................... 39

LIST OF TABLES

TABLE 1 - DESCRIPTIVE STATISTICS &
CODINGS ..................................................................... 23

TABLE 2 - CORRELATION MATRIX FOR RANDOM
SAMPLE ..................................................................... 26

TABLE 3 - CORRELATION MATRIX FOR PRISON-ONLY
SAMPLE ..................................................................... 28

TABLE 4 - LOGISTIC REGRESSION MODEL FOR PRISON
DISPOSITION ............................................................... 31

TABLE 5 - OLS MODEL FOR SENTENCE
LENGTH ..................................................................... 33

vii

INTRODUCTION

In the United States a disproportionate number of African-Americans are housed
in jails and prisons. For example. in 1998 there were 17,362 incarcerated whites as
opposed to 23,586 blacks in the State of Michigan (Bureau ofJustice Statistics, 1998).
What explains this disparity in incarceration practices? Is it a product of systematic racial
bias in the administration ofjustice? Is it a reﬂection ofbiased beliefs on the part of
judges? Or, do African-Americans simply commit more crime? The answer to the
disparity question has been widely debated in the criminological literature. Although
several studies have attempted to determine where the weight of the evidence lies
(Hagan, 1974; Kleck, 1981; Spohn et al., 1981), the issue remains contentious.

Although some observers have concluded that the disproportionate number of
African-Americans in our prisons and jails is a product of unfair sentencing practices that
discriminate against racial minorities (Sellin, 1928; Hagan, 1974; Pope, 1975; Chambliss
& Seidman, 1971; Quinney, 1970; Levin, 1977), others have highlighted wealth as a
more salient factor. More specifically, poor defendants are usually unable to secure
private, competent attomeys and are less likely to gain pretrial release (Burke & Turk,
1974; Lizotte, 1978; Swigert & Farrell, 1977). In the United States, however, race and
wealth often times go hand-in-hand (see Wilson, 1987). So, discrimination based on
wealth is likely to disproportionately affect black defendants because they are more likely
to be poor. Accordingly, wealth discrimination can be viewed as an indirect form of
racial discrimination.

Still, others argue that disparity in sentencing and incarceration stems from

various legal factors, such as seriousness of offense and prior criminal record (Kleck,

1985; Griswold, 1987; Wilbanks, 1987; Kramer & Ulmer, 1996). The argument follows
that since an African-American defendant is more likely to face a serious charge or have
a more extensive prior record, he/she is also more likely to receive a stiffer sentence. It
must be noted, however, that empirical evidence exists to the contrary. In fact, some
research, albeit fairly dated, shows that white defendants receive signiﬁcantly more
punishment for their crimes than blacks (Bernstein et al., 1977; Bullock, 1961; Gibson,
1978). For the most part, however, these studies are treated as anomalies because they
are inconsistent with one of the primary conceptual frameworks driving this line of
research - - conflict theory.

This thesis explores the issue of racial disparity in sentencing using 1998
sentencing data from the State of Michigan. In so doing, two primary questions will be
investigated. Are African-American offenders sentenced to prison more often than
whites? Do African-American offenders who are sentenced to prison receive more
lengthy sentences than white defendants? As noted above, conﬂict theory provides an
analytical framework for better understanding why African-Americans receive more
severe punishment than whites for all crimes, under all conditions, and at similar levels

of disproportion over time (see Hawkins, 1987).
RACE & SENTENCING: A THEORETICAL FRAMEWORK

When conducting scientific research, the investigator must select a theoretical
framework to guide his or her analysis. In the social sciences there are usually a host of
theories to choose from. This is true here. In this section three “theories” (term used

loosely) are identiﬁed and articulated. Each theory provides insight into the controversial

issue of race and sentencing. Ultimately, only one ofthese theories (i.e., conﬂict theory)

will be selected and will serve as the driving force behind the analysis that follows.

The “Overt Racial Discrimination " Hipothesis

One way to explain racial differences in sentencing is to posit that judges use race
as a determining factor when handing down sentences. Walker, Spohn and Delone
(1996) refer to this view as the “Overt Racial Discrimination” hypothesis. It is possible
to couch this perspective in a richer theoretical tradition. It is my contention that the
basic premise of the overt racial discrimination hypothesis is rooted in “labeling theory.”
One of the best-known labeling theorists is Howard Becker. Becker (1963) argued that
individuals, such as criminal justice professionals (e. g., police, prosecutors, and judges),
label people as “offenders.” Once the labeling process is complete, negative outcomes
may follow.

Researchers have presented explanations of how this labeling process might affect
the sentencing decisions of judges. For example, Spohn (1990) used data on the
sentencing practices of judges in Detroit from 1976 through 1978. She found that judge’s
decisions were inﬂuenced most by the seriousness of the crime and by prior criminal
record. It was also noted, however, that the judges included in her study sentenced black
offenders more harshly than whites. In addition, black offenders were more likely to be
incarcerated. Could these last two ﬁndings be the result of judges labeling black
offenders as “more dangerous” and deserving of “harsher sentences?” Goffrnan looked at
issues of identity and of interpersonal relationships that might face a stigmatized
individual. The perceived undesirability of a particular personal property, and its

capacity to trigger stigma-related processes has a history of its own (Goffrnan, 1963).

These personal properties include, but are not limited to. membership in a racial minority
group (also see Schur, 1965).

Although labeling theory appears to lend itselfto explaining the possible mental
state of the judges with respect to their reasoning in sentencing, there are still many other
important aspects of the sentencing process, such as legal variables (i.e., severity of

offense), which labeling theory does not take into account.

The “Minorities Commit More Crime " Thesis

A perspective that does take into account legal variables, such as the severity of
offense and criminal history, is the “Minorities Commit More Crime” thesis. Others,
such as Walker et al. (1996), refer to this approach as the “no racial discrimination”
hypothesis. Here, it is posited that racial and ethnic minorities commit more serious
crimes and have more extensive criminal records; hence they receive more severe
sentences (Hagan, 1974; W ilbanks, 1987; Walker, Spohn & Delone, 1996). It can be
inferred that more severe sentences are imposed on minorities because of legitimate legal
factors rather than the effect of racial prejudice on behalf of judges (Spohn, 1990; Walker
et al., 1996).

Wilbanks (1987) argues that extra-legal variables (e.g., race, sex, age, and the
socioeconomic status of the defendant) do not inﬂuence sentencing decisions (e.g., length
of sentence), but rather legal variables (e.g., type of crime and strength of evidence) are
more important determinants. Hagan (1979) presents an alternative view of sentencing,
which may be referred to also as the “legalistic” viewpoint. The most prominent
variables in Hagan’s model are the defendant’s prior conviction record and the nature and

number of charges brought against the alleged offender (Hagan, 1979). Kramer and

Ulmer (1996) also found that legally prescribed variables, especially criminal history and
offense severity, were strong predictors of dispositional departures (also see Peterson &
Hagan, 1984).

The “Minorities Commit More Crime” thesis ﬂies in the face of empirical reality.
In sum, despite the fact that legal variables do a fairly good job at explaining sentencing
decisions, several studies still show that African-American offenders receive harsher
sentences after legal variables are taken into account (e.g., Lizotte, 1978; Kramer &
Ulmer, 1996; Spohn, Gruhl, & Welch, 1981). In the attempt to take into account for
these differences across racial / minority groups, another theory that provides a more
complete explanation - - conﬂict theory - - is adopted for the study at hand. Put simply,
conﬂict theory addresses more holistically the inter-play between race and disparity in the

sentencing process.

The Conﬂict Perspective

Conﬂict theory is notjust a sociological explanation of criminal behavior, but also
an explanation of the administration of justice. Conﬂict theorists maintain that social
factors explain what behaviors will be defined as criminal, which people commit crime,
and how crimes are investigated, prosecuted and punished. The basic premise of conﬂict
theory is that the law is used to preserve the power of those who possess it, and to control
the behavior of individuals who threaten existing power structures (Quinney, 1970).
Walker et al. (2000) assert that conﬂict theory explains racial disparities in the
administration of justice, and also explains broader patterns of social, economic, and

political inequality in the United States.

Conﬂict theory posits that people of all classes are equally likely to violate the
law, but offenders who occupy lower societal positions are signiﬁcantly more likely to
have harsher sanctions imposed on them (Chambliss & Seidman, 1971). Since racial
minorities are disproportionately found near the bottom of the socioeconomic spectrum,
conﬂict theorists predict that non-white offenders will receive more severe sentences than
whites.

To better understand conﬂict theory, the work of several theorists should be
highlighted. Perhaps the best-known theorist under the conﬂict heading is Karl Marx.
Marx conceptualized social life in economic terms (see Marx & Engles, 1848). Marx
denounced the division of labor in capitalist societies as the unjust exploitation of one
social class (the proletariat) by another (the bourgeoisie). To resolve the exploitation of
the working-class, Marx maintained that capitalist systems must be overthrown. Put
simply, once the “bottom-dwellers” realized that scarce resources were unevenly
distributed, the legitimacy of the state would be questioned. Under such conditions,
revolt and even revolution would occur. After such an event, power and wealth would be
redistributed in a more fair and just manner. Needless to say, the type of revolution
espoused by Marx never occurred in the United States, and the rich kept getting richer
and the poor, as one author noted, “get prison” (Reiman, 1979).

During the Great Depression of the 19305, Thorsten Sellin (193 8) argued for a
more broad deﬁnition of “crime” and a less conventional approach to criminological
theory. Sellin (1938) argued that “culture conﬂict” is a source of crime. He explained
that different groups learned different “conduct norms” and that the conduct norms of one

group sometimes clashed with those of other groups. Sellin (193 8) deﬁned “conduct

norms” as cultural rules that require certain types of people to act in certain ways under
certain circumstances. Sellin called this phenomenon “primary cultural conﬂict.” With
respect to which conduct norms would inﬂuence the formulation of criminal law, Sellin
(1938:107) noted “the conduct which the state denotes as criminals, of course, that
deemed injurious to society, or in the last analysis, to those who wield the political power
within that society and therefore control the legislative, judicial and executive functions
which are the external manifestations of authority.” Vold and Barnard (1986) state that
in more complex societies overlap and contradiction exists between the conduct norms of
different cultural groups. Sellin (193 8) also noted that a single culture could evolve into
several different sub-cultures, each having its own set of conduct norms, which can also
conﬂict with one another (termed “secondary culture conﬂict”). Sellin’s work suggests,
then, that substantive criminal law would not represent a consensus view, but rather
would reﬂect the conduct norms of the dominant culture.

George Simmel, a German scholar with a deep interest in social theory, proposed
“sociological formalism” (see Martindale, 1960). While Simmel was interested in
conﬂict as a common feature of social life, he was not preoccupied with it as was Marx.
Lilly, Cullen & Ball (1989) note that Simmel saw conﬂict not as a problem necessarily
calling for solutions or even leading to change, but rather as a typical feature of social
order. In Simmel’s view, conﬂict actually contributed to social order. Simmel also
differed from Marx in that he did not focus on the causes of conﬂict, but instead focused
on the consequences of conﬂict.

In the 19503, Vold (1958) presented an extensive and detailed treatment of

criminological theory from the conﬂict perspective. He argued that as social interaction

“processes grind their way through varying kinds of uneasy adjustment to a more or less
stable equilibrium of balanced forces in Opposition, the resulting condition of relative
stability is what is usually called social order or social organization” (Vold, 1958:204).
This inter-group conﬂict was conceptualized in Simmelian terms as potentially making a
positive contribution to the strengthening of different groups because participation in the
struggle resulted in group esprit de corps and solidarity (Lilly, Cullen & Ball, 1989).
Vold posited that it was normal for groups in a complex society to come into conﬂict as
their interest collided and that “politics . . . is primarily a matter of ﬁnding practical
compromises between antagonistic groups in the community at large” (Vold, 1958: 208).
Vold (1958: 209) refers to legal compromises cited by Edwin Sutherland: “those who
produce legislative majorities win control over the police power and dominate the
policies that decide who is likely to be involved in violation of the law.” Vold considered
politics the art of compromise and noted “ the principle of compromise from positions of
strength operates at every stage of this conﬂict process” (Vold, 1958: 209).

Vold posited many crimes were of a political nature. He alludes to crimes
resulting from protest aimed at political reformation in that “a successful revolution
makes criminals out of the govemment ofﬁcials previously in power, and an unsuccessful
revolution makes its leaders into traitors subject to immediate execution” (Vold, 1958:
214). Vold asserted “numerous kinds of crimes result from the clashes incidental to
attempts to change, or to upset the caste system of racial segregation in various parts of
the world notably in the United States and in the union of South Africa” (Vold,

19582217).

Richard Quinney’s (1970) six propositions ofthe “social reality ofcrime” are also
informative. For the purposes ofthis review, only four ofhis propositions will be
discussed. Quinney (1970:15) posited “crime is a deﬁnition of human conduct that is
created by authorized agents in a politically organized society. Here conﬂict and labeling
theory inter-mingle in that those who possess power and legal authority deﬁne what is
criminal. He also maintained “criminal deﬁnitions describe behaviors that conﬂict with
the interests of segments of society that have the power to shape public policy”
(1970:16). Once again, conﬂict and labeling perspectives merge as the author posits
those in power label what is criminal and then make laws to address the behaviors
deﬁned as such. Quinney (1970: 1 8) also posited that “criminal deﬁnitions are applied
by the segments of society that have the power to shape the enforcement and
administration of criminal law.” He asserted that the “probability that criminal
deﬁnitions will be applied is inﬂuenced by such community and organizational factors as
(1) community expectations of law enforcement and administration, (2) the visibility and
public reporting of offences and (3) the occupational organization, ideology and actions
of the legal agents to whom authority to enforce criminal law is delegated” (Quinney,
1970:19-20). Finally, Quinney (1970220) argued, “behavior patterns are structured in
segmentally organized society in relation to criminal deﬁnitions, and within this context
persons engage in actions that have relative probabilities of being deﬁned as criminal.”
Here the author asserts that those actors not in power are excluded from the policy
formulation process, hence there is a greater chance their behavior will be deﬁned as

criminal more often.

Conventional criminology has provided the knowledge necessary in detecting and
controlling those groups and individuals perceived as threats to the powerful. Platt
(1969) contends that the original establishment of the juvenile court was not a means of
helping youth but rather a technique by which upper-class, Republican, protestant women
extended control over the children of the Catholic and Jewish immigrants. In a similar
vein, the concept of rehabilitation was also criticized as a tool of political oppression
which justiﬁed prolonged and invasive tinkering with the minds and bodies of prisoners
under the guise of assisting them (Smith & Fried, 1974). Conﬂict theory has succeeded
in providing for a broadened reorientation and an increased sensitivity to issues

previously overlooked or threatened only in passing (Thomas & Hepburn, 1983).
REVIEW OF THE RESEARCH LITERATURE

In this section, a detailed review of the literature organized around three
theoretical frameworks identiﬁed above will be provided. Doing so will provide
justiﬁcation for the variables that will be selected in the analysis presented below.

As previously noted, labeling theory has also been used as an avenue of
explanation for sentencing disparity between whites and non-whites - - the “Overt Racial
Discrimination” hypothesis. Scholars have argued that the most important factor
regarding disposition is the degree to which formal social control organizations (e. g.,
police, prosecutors, and judges) determine which individuals will be deﬁned as deviant
(Chambliss & Leill, 1966). With regards to plea bargains, constitutionally-based
arguments against plea bargaining have focused largely on the issue that defendants who

exercise constitutionally guaranteed fair—trial rights are penalized for doing so (Uhlman &

10

Walker, 1979). Could this be interpreted as meaning that going to trial and pleading “not
guilty” when there is evidence to the contrary in some way stigmatizes the defendant as a
person who is not remorseful for what he or she has done? Could this perceived lack of
remorse yield harsher sentences? Miethe and Moore (1986) found race to be a signiﬁcant
predictor of sentence length even when legal controls were introduced. The authors’
race-speciﬁc model showed that blacks, especially those who were deemed high-risk
felons, were not married, did not complete high school, were unemployed, used a weapon
in the commission of their criminal offence, and had prior felony convictions, were
treated more harshly. In contrast, black offenders who were classiﬁed as low-risk felons,
were married, had completed high school, were employed, did not use a weapon, and did
not have a felony record, were not treated more harshly.

Burke and Turk (1974) found that men who have been institutionalized prior to
their most recent conviction were signiﬁcantly more likely to be brought to court,
convicted, and given a prison sentence instead of a less severe punishment after
controlling for race and other legal variables. With regards to juvenile offenders, the
authors note that the home situation of nonwhite youths is “perceived” by authorities as
less adequate than that of whites, which results in nonwhites being institutionalized at
higher rates and white juveniles being released to the custody of their parents (Burke &
Turk, 1974).

With regards to judges sentencing decisions, Kramer and Ulmer (1996) found
race to be a signiﬁcant predictor of dispositional departure decisions even when legal
controls were introduced. When judges are confronted with a dispositional departure

decision, they may not only rely on legal variables (e. g., prior convictions and offense

ll

 

severity), but also on the defendant’s plea decision, as well as extra-legal variables (e.g.,
race, sex, and age) (Kramer & Ulmer, 1996). The latter information may be used to
stereotype defendants. Using these stereotypes, court actors may project behavioral
expectations about whether the offender will beneﬁt from rehabilitation programs,
whether the offender is a potential danger to the community, and whether the offender is
more or less blameworthy, and thus more or less deserving of punishment (Stone, 1962).
Peterson and Hagan’s (1984) study of sentencing in drug-related cases was
conducted during the late-19605 and 1970s. During this time, the Nixon Administration

reformed federal drug statutes. The authors contend that the American drug prohibition

rW“!

began with the portrayal of minorities as the villains responsible for the growing drug
menace (Peterson & Hagan, 1984: 67). Finally, Albonetti (1991: 261) found that black
defendants who entered the sentencing stage early were deﬁned as “posing a danger to
the community” or a “risk of failure to return for pretrial processing” - - a legally relevant
criterion at the bail hearing in the District of Columbia - - which indicates that under
conditions of heightened uncertainty (deﬁned by bail amounts) judges are more likely to
rely on stereotypes of black defendants as being more likely to recidivate. As a result,
judges impose harsher sentences on minorities relative to white defendants who have the
same ﬁnancial bail conditions (Albonetti, 1991). This section denotes primarily that it is I”~
the perception of those in power towards minority races as “dangerous,” which frequently
manifests itself through more harsh sentencing outcomes for minority defendants.

The “minorities commit more crime” thesis has its fair share of supporters.
Blumstein (1982) concludes that the differential involvement of blacks as arrestees,

especially for the offenses of homicide and robbery, accounts for 80% of the disparity

l2

between black and white incarceration rates. Furthennore, Blumstein (1982) found that
as the seriousness of the offense decreased, blacks were disproportionately represented in
prison. It appears, then, that as the severity of the crime decreases judges use more
discretion when rendering their sentencing decisions.

Hagan’s (1974) “legalistic” approach, which de-emphasizes extra-legal attributes
(e.g., race, sex, age and socioeconomic status of the alleged offender), focuses on prior
convictions as well as the nature and number of the charges brought against the
defendant. The “legalistic” approach closely parallels the “minorities commit more
crime” thesis. Overall, the research in this area suggests that a modest relationship (at
best) exists between race and disposition once legalistic variables are included in the
equation. Pratt‘s (1998) meta-analysis of forty-seven studies is informative with regards
to understanding disparity in sentencing. Pratt (1998) found that offense severity was the
only signiﬁcant variable related to the length of sentence. He noted also that neither race
nor prior criminal record variables were found to be signiﬁcant covariates of sentence
length. Pratt (1998: 520) cautions, however, that researchers investigating the differential
treatment of racial minorities in the criminal justice system would be well advised to pay
special attention to measurement issues because they can inﬂuence research ﬁndings.

Spohn (1990) notes that judges sentencing decisions are determined most by the
seriousness of the crime committed and by the offender’s prior criminal record. Spohn
(1990) found that offenders convicted of more serious crimes or a greater number of
charges were punished more harshly as were offenders who used a gun, inﬂicted harm on

their victim(s), or victimized a stranger. After investigating the matter further, Spohn

13

 

 

(1990) found that the race ofthejudge had little inﬂuence on case disposition and was
overshadowed by legal factors.

In his study of“high control jurisdictions”(i.e., El Paso, New Orleans & Seattle),
LaFree (1985) found that the race of the defendant was a signiﬁcant predictor of sentence
outcomes, but the magnitude of the effect was relatively modest. More speciﬁcally,
LaFree’s (1985: 298-306) results suggest that ten other variables (e.g., defendant’s
criminal record, whether the offender was probation or parole, whether a weapon was
used) were more important than the race of the defendant. In “low control jurisdictions”
(i.e., Pima County-Tucson, Norfolk, and two counties in Pennsylvania), LaFree (1985:
298-306) found that the best predictors of sentence severity were measures of statutory
seriousness and evidence - - not the defendant’s race.

Another variant of the “Minorities Commit More Crime” thesis is the “differential
involvement” perspective. The differential involvement perspective holds that racial
minorities, particularly African-Americans, are over-represented in prison populations
and are given harsher sentences than whites because they are involved in the commission
of more crimes (Wilbanks, 1987; Wilson & Hemstein. 1985). This view contends that
African Americans receive longer sentences because, on average, their crimes tend to be
more serious and thus warrant harsher punishment (Hagan & Albonetti, 1982; Hindelang,
1969; Kleck, 1981, 1985; Peterson & Hagan, 1984; Wilbanks, 1987b).

The conﬂict perspective has also guided criminal court processing research.
Researchers using this theoretical framework hypothesize that defendants at the bottom
of the social heap will receive more severe dispositions (Unnever, Frazier & Henretta,

1980). Simply put, conﬂict theorists posit that the law is applied differentially to protect

l4

the hegemony ofa ruling elite. Citizens at the top ofthe social strata do violate the law,
but they are less likely to be held accountable because they have the power and resources
to avoid severe dispositions. Sellin (1928) noted that some groups (i.e., poor citizens)
lack political and social status and are more likely to suffer hardships when dealing with
the police and courts. In sum. the administration of criminal justice fails those at the
margins of society by not providing necessary safeguards against arrest, trial and
conviction.

The differential treatment of blacks may also be seen prior to sentencing in the
criminal courts. For example, Zatz and Hagan (1985) argue that determinant sentencing
may encourage prosecutors to exercise more discretion in deducing which cases are
worthy of full prosecution. Conﬂict theorists contend that prosecutors who are mainly
white and representative of the ruling class will exercise their legal authority
disproportionately on minority defendants.

Other factors, such as the race of the victim and the race of the defendant, may
also be interpreted within the conﬂict theoretical framework. For example, Quinney
( 1970) posits that courts tend to regard the slaying of a white by an African-American as
evidence of guilt. The same cannot be said, according to Quinney, if the alleged offender
was white and the murdered victim African-American. Does this dynamic reﬂect the
power in-balance between the “haves” and the “have-nots?”

Several authors have found race to be a signiﬁcant factor in sentencing decisions
even after legal factors were introduced into the equation (Spohn, Gruhl & Welch, 1981;
Levin, 1977; Pope, 1975; Unnever, Frazier & Henretta, 1980). Pope’s (1975) study on

Superior Court sentencing in rural areas, which included both jail and prison

15

 

commitments, found that 66% of all white offenders were sentenced to terms of
incarceration compared to 79% of black offenders - - a 13% difference (Pope, 1975).
Unnever, Frazier and Henretta ( 1980) found that even when controlling for important
legal and extra-legal variables, there was a direct race effect. Put differently, the authors
found that whites had an 18% greater chance of receiving probation than blacks when all
else was equal (Unnever, Frazier & Henretta, 1980). Spohn, Gruhl and Welch, (1981)
hypothesized they would ﬁnd no direct racial discrimination either in the decision to
impose a more or less severe sentence or in the decision to inearcerate. Their hypothesis
was not conﬁrmed. Even after controlling for both legal and extra-legal factors, black
males still were sentenced to prison more often than white males. Levin’s (1977)
analysis of sentencing decisions in Pittsburgh and Minneapolis found that whites received
shorter terms of incarceration than blacks (Levin, 1977). These ﬁndings lie in stark
contrast to studies reporting that race is not a signiﬁcant predictor once legal variables
(i.e., prior convictions and offense severity) are taken into account.

Research also indicates that the association between the defendant’s race and the
outcome of the case may be inﬂuenced by social class (Swigert & Farrell, 1976, 1977).
Accordingly, it has been argued that blacks are more severely sanctioned because they -,_
are disproportionately members of the lower class. Lizotte (1978) studied the indirect
effects of occupation and race on prison sentence length. Not only did Lizotte (1978)
ﬁnd that non-whites and people with lower occupational prestige were less likely to make
bail, but he also found these factors also inﬂuenced the length of prison sentences.
Bridges and Crutchﬁeld (1988) found that social characteristics of states contribute

signiﬁcantly to racial disparity in imprisonment. The social standing of blacks when

16

compared to whites fosters disparity in imprisonment by increasing the chances of blacks
being imprisoned and decreasing the chances of whites being imprisoned, even after
racial differences at arrest are considered (Bridges & Crutchﬁeld, 1988). The lower
class defendants who possess little status in the community, low occupational prestige,
and a personal life frequently described as disorganized enters the court process with
little to lose (Swigert & Farrell, 1977). More afﬂuent citizens, on the other hand, have
suffered greatly simply because they have been arrested. Suspended professional license,
loss of status in the community, personal and familial trauma is seen as being a sufﬁcient
punishment (Swigert & Farrell, 1977). This review of the literature has provided an in-
depth discussion of the causes and circumstances that lead to the overrepresentation of
racial and ethnic minorities in American correctional institutions. While some
controversy persists in the research literature, it is the author’s view that institutionalized
racism manifested covertly through the American criminal justice system is the most

signiﬁcant determinant of racial disparity in the sentencing stage of the court process.
RESEARCH OBJECTIVE

Conﬂict theorists argue the application of criminal sanctions will vary according
to the extent to which the “behavior of the’powerless conﬂict with the interest of the
power segments” (Quinney, 1970218). However, some authors still argue that conﬂict
theory needs to be re-conceptualized (Hawkins, 1987; Peterson & Hagan, 1984). For
example, a fair amount of evidence suggests that judges’ decisions are inﬂuenced by the
seriousness of the crime, the weight of the evidence, as well as other legal factors (e. g.,

the number of charges and the defendant’s prior criminal record). A rigorous test of

17

conﬂict theory with regards to sentencing outcomes would take these legal factors into
account. Therefore, racial discrimination will be deﬁned as disparity in the sentencing
process between members of different racial groups once legal variables are taken into
account. Accordingly, the following questions are proposed for empirical inquiry. Are
racial minorities more likely to be sentenced to prison than whites once legal factors are
taken into account? Are racial minorities more likely to receive longer sentences? To
investigate these questions, a sample of 1998 district court dispositions from the State of

Michigan are used to estimate a series of multivariate models.

rim-m

METHODS
Data
The original data set includes offender-based information, such as demographic

characteristics, offense information, and case outcomes, from 83 district courts in the

. State of Michigan for the year 1998. These data were collected by the Michigan

Department of Corrections (MDOC). The original data ﬁle contains a total of 56,844

sentencing dispositions. For the purpose of the analysis two samples were taken from the

original data ﬁle. First, a random sample of 8,000 sentencing dispositions was drawn

from the original total population ﬁle. A random sample was taken because such I
i

techniques ensure that the sample will be highly representative of the population. The

 

results provided in Table 1 support this contention. The second analysis ﬁle was created
using the random sample, and only contains offenders who were sentenced to prison (N =
2,065)(referred to hereafter as “prison-only sample”).

Overall, these data provide a wealth of information, and will allow for the

operationalization of various independent and dependent variables. There are, however, a

18

few limitations that should be noted. First, the data ﬁle does not provide infomiation
regarding the characteristics ofthe victims (ifthere was a victim). It is also impossible to
determine the type of plea entered by the offender (e.g., “guilty”). Lastly, in the event of
a trial, it is not possible to determine whether the offender selected a “bench” or “jury”
trial. Despite these shortcomings, the data ﬁle provided by the MDOC provides a wealth

of information that is well suited for the research objective at hand.

Dependent Variables

For the purpose of determining whether sentencing disparities exist with regards
to the defendant’s race, two dependent variables have been selected. The ﬁrst variable
(termed “sentenced to prison”) is a dummy variable indicating whether the offender’s
sentence included incarceration in a state prison (1 = yes, 0 = otherwise). This variable
will be used to determine whether black offenders were more likely to receive prison
sentences when compared to white defendants. As can be seen in Table 1, 25 percent of
all offenders were sentenced to prison in 1998. The second outcome variable, termed
“sentence length” is an interval-level measure reﬂecting the maximum term imposed (in
years). This dependent variable will be used to address the question of whether blacks
were signiﬁcantly more likely to be sentenced to longer durations than white defendants.
The mean number of years is 11 within the prison-only sample. This variable is
positively skewed with the majority of offenders receiving maximum sentence lengths of

20 years or less.

19

Independent Variables

Sixteen independent variables will be used in the analyses. These variables can
be grouped into three categories: offender characteristics (or extra-legal variables), legal
variables, and control variables.

Offender Characteristics. The following analyses will include ﬁve offender
characteristics. The ﬁrst variable, gender, is a dummy variables which is coded 1 = male
and 0 = female. Males comprised 85 percent of the random sample, and 93 percent of the
prison-only sample. This variable is included because departure decisions may be
inﬂuenced by the defendant’s gender. The age variable is continuous and reﬂects the age
of the convicted offender in years. The average age for offenders in the random sample
was 30 years, and 30 years in the prison-only sample. Finally, three race / ethnicity
dummy variables are also included: black, Hispanic, and “other race.” For each of these
variables, Caucasians are used as the reference group. Blacks comprised 40 percent of
the random sample and 45 percent of the prison-only sample. Hispanics and other
members of other racial minority groups were less well represented in both samples (see
Table 1). Previous studies using multivariate modeling techniques have included
offender characteristics (see, for example, Unnever, Frazier & Henretta, 1980; Miethe &
Moore, 1986; Swigert & Farrell, 1977; Blumstein, 1982).

Legal Variables. Ten legal variables are also included in the analyses. The ﬁrst
variable, ﬁrearm possession, is a dummy variable indicating whether the offender
violated any existing gun laws during the time of their offense (1 = yes, 0 = otherwise).
Zatz and Hagan (1985) found the use of a ﬁrearm increased sentence length. The

offender’s legal status at the time of the offense is also included in the following

20

analyses. Doing so is accomplished by using a series ofdummy variables: out on bond
(1 = yes, 0 = otherwise), parolee (1 = yes, 0 = otherwise), probationer (1 = yes. 0 =
otherwise,), or incarcerated offender ( l = yes, 0 = otherwise.) Offenders not falling into
one of the four predetermined categories (e. g.. citizens not under correctional supervision
or on pretrial release) are used as the reference category.

Criminal activity will be gauged using variables under two broad categories.
First, prior incarceration will be captured using two variables: number of times the
offender has been incarcerated in jail (prior jail) or prison (prior prison; 1 = yes, 0 =
otherwise). Existing research shows that prior jail and prison terms inﬂuence sentencing
decisions (Pope, 1975). Second, the nature of the offense is captured using three dummy
variables: drug offender, violent offender, and property offender. Other offenses, such
as public order violations, are used as the excluded category. Previous studies have
found the severity of the present offense to be a signiﬁcant predictor of sentencing
decisions (Spohn, 1990; Pratt, 1998; Uhlman & Walker, 1979; Blumstein, 1982; Zatz &
Hagan, 1985; Peterson & Hagan, 1984).

Control Variable. To further guard against spurious ﬁndings within the statistical
analysis, an additional dummy variable will be used. Previous studies have controlled for
pre-sentence investigation (PSI) reports (1 = yes, 0 = othem'ise)(Peterson & Hagan,
1984; Unnever, Frazier & Henretta, 1980). Unnever, Frazier & Henretta (1980: 205)
noted that PSI’s might inﬂuence sentencing outcomes because race bias may subtly enter
the criminal justice process and be passed on in the form of sentencing recommendations.

Descriptive statistics for the variables used in the analysis are provided in Table l.

21

Statistical Procedure

Two different multivariate statistical models will be used in the following
analyses. To assess the effects of the independent variables on whether the defendant
received a sentence that included incarceration, which is binary response variable,
logistic regression will be used. Logistic regression is an appropriate technique when the
analyst is interested in regressing a dummy variable on a set of independent variables
with different levels of measurement (Bachman & Paternoster, 1997). For the analysis
regarding length of imprisonment, Ordinary Least-Squares (OLS) regression will be
employed. OLS is a multivariate technique that is designed for modeling the effects of
independent variables on a dependent variable that is interval in nature. Because
collinearity concerns often arise when estimating multivariate models, tolerance tests will
be used to assess the strength of the linear relationships between the independent
variables. In addition to the above techniques, bivariate techniques will also be used to

diagnose potential statistical problems, as well as to conduct preliminary hypothesis tests.

22

Table 1

Descriptive Statistics (Q Ceilings
Population

Variables

Sentenced to Prison

1: Yes

0: Otherwise
Sentence Length

it Of Years
Black

1= Yes

0: Otherwise
Hispanic

1= Yes

0= Otherwise
Other Race

l= Yes

0= Otherwise
Gender

1= Male

0: Female
Age

it Of Years
Firearm Possession

l= Yes

0= Otherwise
Out on Bond

1= Yes

0= Otherwise
Parolee

1= Yes

0= Otherwise
Probationcr

1= Yes

0= Otherwise
Incarcerated
Offender

l= Yes

0= Otherwise
Prior Prison

1= Yes

0: Otherwise
Prior Jail

it of lncarcerations
Drug Offender

l= Yes

0: Otherwise
Violent Offender

1= Yes

0: Otherwise
Property Offender

1: Yes

0: Otherwise
Pre- Sentence
Investigation

l= Yes

0= Otherwise

N=

Mean

SD.

Random Sample

Mean SD.

Prison-Only Sample

Mean 8.1).

 

.25

5.34

.41

.84

30.17

.01

.06

.01

.99

56.844

.43

12.06

.49

10.41

.41

.40

.30

.07

.25 .43

5.46 12.34

.40 .49

.01 .10

.84 .35

30.13 10.55

.06 .24

.05 .22

.01 .10

.21 .41

.99 .07

8.000

23

1 .00 .00

11.31 12.33

.45 .49

.02 .15

.00 .09

.93 .25

30.49 10.12

.05 .22

.07 .26

.27 .44

.44 .49

2.12 2.47

.99 .02

2,065

IY-Mﬂ

RESULTS

Bivariatc Results

Bivariatc associations for the variables used in the analysis are presented in
Tables 2 and 3. Not only do bivariate techniques allow the analyst to assess the
correlation between two variables, but they can also be used to determine whether
collinearity problems may present themselves in multivariate analyses. The optimal
situation would be for the independent variables to strongly correlate with the dependent
variable, but only weakly correlate with each other. The results from the correlation
matrices provided in Table 2 and 3 show that none of the coefﬁcients for the independent
variables exceed 0.50. Accordingly, it is reasonably safe to assume that collinearity will
not be an issue in the multivariate analysis. The bivariate results presented here serve
another purpose as well: the results provide a ﬁrst opportunity to assess the hypothesized
relationships. I will ﬁrst discuss the bivariate results assessing whether blacks and other
racial and ethnic minorities are disproportionately sentenced to prison. Following this
discussion, I will focus on the question of whether blacks are sentenced to longer periods
of incarceration in state prisons.

Table 2 presents the bivariate relationships for the variables used in the T
“sentenced to prison” analysis. Again, theiprimary question concerns whether African-
American offenders are sentenced to prison more often than whites. A positive
association was observed between whether the offender was sentenced to prison and
whether the offender was black (r = 0.06). Although statistically signiﬁcant, the
magnitude of the association was very weak. In other words, the hypothesis that one’s

race inﬂuenced sentencing decisions in Michigan does not appear to be strongly

24

supported in this analysis. This ﬁnding does not square with Pope’s (1975) contention
that the disproportionate number of blacks in American prisons is a product of unfair
sentencing practices. Similar associations were observed between being sentenced to
prison and the other racial and ethnic groups, such as Hispanic and “other” minorities.
Before proceeding to the question concerning length of imprisonment, one additional
note is necessary. The results provided in Table 2 indicate that blacks are
disproportionately involved in drug offenses (r = 0.24). This ﬁnding is consistent with
prior research (Peterson & Hagan, 1984: 67).

Concerning the other independent variables, weak correlations (r < 0.10) were

. r— ____.. _._. __r..
i I

observed between being sentenced to prison and several variables, including whether an
offender was out on bond or a probationer at the time of their present offense, whether
the offender was guilty of drug or property crimes, whether 3 PSI was made available to
the court, as well as the offender’s age. In relative terms, stronger correlations were

1 observed between “sentenced to prison” and seven other variables: offenders with prior
prison (r = 0.33) and prior jail experience (r = 0.19), offenders who were out on parole at
the time of their offense (r = 0.24) or who were incarcerated at the time of their present
offense (r = 0.14), offenders guilty of ﬁrearm possession (r = 0.19), offenders guilty of
violent crimes (r = 0.17), as well as an offenders gender (r = 0.13). All ofthese

relationships were in the expected direction.

25

._o.VQ_.:._ .mo.va....

 

 

a cor KO. 8. S. .No. cc. :mo. 3. Na. oo. car 8. co. 8. .No. .55. ~wa— .5—
8023—..0
— :hor :n _ .- one: :2..- —¢.- co. oar _o.- :27 :m — .- :MO. 2.. cc. :6 — .- senor mtg—P:— @—
hO‘ﬁh—O
— :m _ .- :mor oar oo. _o.- co. co.- .ea _ . «or «one. .o. co. zoo. 2.: . «Ev—or? .m—
h0g8
— snow «one. cc. cor :vc. NO. oomcr :hc. co. conq- oor :VN. co. HEA— .v—
— :2”. «two. :_ _. .8: . gone. :mor :VN. :56. co. 3. «Nov 3¢ _. al.. hot.— .n—
n no: . :vcr oomv. cc. :vo. :nN. :2. nNor .No.- :v—. :nm. Eva—h.— hctm .N—
hO-v-huuc
— :nor N9- :Nor :vc. _o.- :mc. oer cor «No. :1. 3:09—8:— .=
u :N _ .. 32.. GD. :27 :mc. 2v.- .o. :No.- «oao. hocenuI-Em A:
n :90.- _o. :50. «Lb. co. ~o.- «MO. :VN. 00—5.— d
E
— cor 3.- co. cc. co. one. .NO. 8 «BC d
£33680.—
— comer .36. 8.. oo.- 25C. :2. E...— .6
_ :3- 3.. 8.- S. 8. 0:. .e
— cor one. :nor :2 . .3300 .m
00¢“
— 8... :2..- Sr 50.8 .9
_ :m _ .- 8. 2.1:: .n
n cove. dun—ﬂ .N
ﬂan—ha— 6h.
— House—now .—
h— on m— 9— Mn Nu = A: a a b w n v n N —

 

Goo .w H75 335%. Sonnet Lo\k.5§< $233.89

N nigh.

26

Table 3 presents the bivariate relationships between the variables used in the
“sentence length” analysis. The primary question here concerns whether African-
American offenders receive harsher (i.e., longer) prison sentences than whites. A
positive association was observed between sentence length and whether the offender was
black (r = 0.08). Although the coefﬁcient is statistically signiﬁcant, the strength of the
association is very weak. Subsequently, the hypothesis that one’s race inﬂuences
sentence length in Michigan does not appear to be strongly supported in this analysis.
This ﬁnding is consistent with Wilbanks (1987) contention that extra-legal variables,
such as race, do not inﬂuence the length of sentence dispositions. Similar associations
were observed between sentence length and the other racial and ethnic groups, such as
Hispanic and “other” minorities.

Concerning the other independent variables, weak correlations (r < 0.10) were
observed between sentence length and several variables, including gender, age, out on
bond, parolee, probationer, incarcerated offender, prior prison, drug offender, as well as
PSI. The associations between sentence length and four other variables were more
notable. Property offenders and those with prior jail experience were incarcerated for
shorter durations. Violent offenders, in contrast, received longer sentences. One of the
most impressive correlations, at least with regards to the strength of association, was
between sentence length and ﬁrearm possession (r = 0.31). This ﬁnding indicates that
offenders who were in possession of a ﬁrearm at the time of the crime received more

time.

27

 

 

8.x? wove...

 

 

3. CO. :v. nor cc. 3. co. oo. as. co.. co; co. co. «6.. oc.- ~w& .h—
hO‘ﬂOh—O

:2. :2- 8. 8.- 8.- 8. 8. 8.- :8.- ..8.- 8. 8.- 8.- :8.- ..2 .. Exp... .2
LO‘p-Og

— scour :N~.. 992%. No.- nor «an?- _o.- :VN. coho; :oo. onc. No. coac. ¢.:- «HO—OS .mu
50308

— mcr No. 3.- cc.- .ino. oo. :56.- S. 8.. No.. 8.. :nn. coho. "EA— .9-

— :wu. no. «we. :2. no. :2.- ..mm. No; 8.. no. .cwor .83.- =-—. her-A— dun

u .9?—. con—r conv. nor .vcr :MN. no. 8.. 8.. :2 . one: gar-A— hO-hh .N—
segue

— ecu—r :woc .09. S. .95.. ma. 3. no: N0. N9. 13:00.30:— .m—

— :nNr poh _ .- 8.. one _ - :27 co.. No. 92...- _o.- bonsai—ah .6—

— :_ _.. mar :oo. :09. no. 8.- :wo. cc: oo—P-Im .a

— Nor evcr cc. 2.. 3. Mc- eowcr Hg 5 «HO d

u :59. «owe. N66 N9. «05— . a: n. 5.9:...— .h

_ :2 .- 8.- 8. .8. 8. «3 .o

— co. cor 8.. VG. hog d

u 3.. :wcr 2H. 5 52—5 .9

_ :n _ .- 8. 9.53: .n

— nowo. in .N
59.3

n 0063:“ .—

m— ’4 n— N— = e— a a h w n v n N —

 

A88.” "8 sales-n 85 28.5 SR 5%: 22338.6
8 oz:-

28

To summarize, two important ﬁndings were revealed in the bivariate analyses.
First, the individual’s race and/or ethnicity did not appear to inﬂuence whether the
offender received a prison sentence. nor did it inﬂuence the length of incarceration for
those offenders who were sent to prison. Second, variables falling under the “legalistic”
heading were stronger correlates of sentence disposition (i.e., whether sent to prison) and
sentence length. In short, the bivariate analyses indicate that race does not appear to
matter much in the State of Michigan. However, before more deﬁnitive conclusions can
be drawn, more rigorous assessments of the hypotheses of interest need to be conducted.
Toward this end, multivariate analyses ( e. g., OLS and logistic regression) are presented
below that tests which factors (e. g., legal and extra-legal) predict sentencing outcomes in

Michigan for the 1998 calendar year.

Multivariate Results

Some scholars argue that disparity in sentencing and incarceration stems from
various legal factors, such as the seriousness of the offense and prior criminal record
(Kramer & Ulmer, 1996). Still, others contend that the disproportionate number of
African-Americans in our prisons and jails is a product of unfair sentencing practices that
discriminate against racial minorities (Quinney, 1970). Tables 4 and 5 show the results
of the multivariate analyses. Speciﬁcally, Table 4 provides the results for a logistic
regression model predicting whether the offender was sentenced to prison. Table 5,
which reports the results from an ordinary least-squares regression model, reveals the
factors that contribute to sentence length for those offenders who were shipped off the

state prisons. Before proceeding, it should be noted that tolerance tests for all of the

29

independent variables included in the following statistical models showed that
collinearity was not a matter of concern.

Table 4 addresses the question of whether African-Americans are more likely to
be sentenced to prison than whites, after taking into account a host of relevant factors.
Counter to the hypothesis speciﬁed above, the results show that black offenders were not
signiﬁcantly more likely to be sentenced to prison (t-ratio = -1.04). This ﬁnding is
inconsistent with LaFree (1985) who found that the race of the defendant was a
signiﬁcant predictor of sentencing outcomes. It is consistent, however, with others who
have found that the race of the offender is not a statistically signiﬁcant determinant of
sentencing outcomes once legal variables are taken into account (Spohn, 1990; Pratt,
1998; Hagan, 1974). Null effects were also found for Hispanics (t-ratio = 0.24) and
“other” minorities (t-ratio = -1.46) with regards to being sentenced to prison.

Three additional independent variables did not achieve statistical signiﬁcance (p <
0.05). In particular, whether the offender was convicted of a property offense, a drug
offense, or had 3 PSI did not inﬂuence whether they were sentenced to prison. As for the
remaining independent variables, several signiﬁcant ﬁndings emerged. Males and
younger offenders were signiﬁcantly more likely to be sentenced to prison, as were
offenders who were out on bond, on parole, on probation, or who were incarcerated at the
time of the offense in question. Further, prior incarcerations (i.e., prior jail and prior
prison) and offense type (i.e., violent offenders) were signiﬁcant (p < 0.01) predictors of
whether an offender was sentenced to prison. Before proceeding to the question
concerning length of imprisonment, one additional ﬁnding needs to be noted. Offenders

who were guilty of ﬁrearm possession at the time of their present offense were

30

signiﬁcantly more likely to be sentenced to prison. The odd-ratio revealed that these
offenders were thirty-nine times more likely to be sentenced to prison than offenders who
did not use a ﬁrearm during the commission of their crime. This ﬁnding is consistent
with prior research that indicates other variables (e.g., whether a ﬁrearm was used) were
more important predictors of sentence outcomes than the race of the offender (Spohn,

1990; Hagan, 1974; LaFree, 1985: 298-306).

 

 

 

Table 4
Logistic Regression Mode/for Prison Disposition (N = 8,000)

Coefficient SE t-Ratio
Constant -4.26 1.00 -4.26**
Black -.06 .06 -1.04
Hispanic .04 .18 .28
Other Race -.41 .29 -1.46
Gender .74 .10 7.02**
Age -.01 .00 -4.69**
Firearm Possession 3.67 .42 14.64”
Out On Bond .49 .1 l 4.26“
Parolee 1.36 .12 13.06**
Probationcr .75 .06 11.1 1**
Incarcerated 2.22 .28 10.30**
Offender
Prior Prison 1.31 .07 19.02”
Prior Jail .09 .01 7.70**
Drug Offender .16 .08 1.89
Violent Offender 1.21 .08 1439’”
Property Offender -.06 ‘ .09 -.60
PS1 1.94 1.00 1.90
Model Chi-Square 1719.39”
Nagelkerke R-Square .28

 

*p < .05, **p < .01.

Table 5 addresses the question of whether African-Americans receive harsher

prison sentences (i.e., more time) than white offenders after considering several pertinent

31

factors. The hypothesis speciﬁed above is not supported in the regression analysis. The
results show that African-Americans were not signiﬁcantly more likely to receive longer
prison sentences (t-ratio = 0.53). This ﬁnding is consistent with prior research that de-
emphasized extra-legal characteristics (e.g., race and sex) as signiﬁcant predictors of
sentence length (Hagan, 1974; cf. Kramer & Ulmer, 1996). Null effects were found as
well for Hispanics (t-ratio = 0.86) and “other” minorities (t-ratio = -0.17) with respect to
receiving harsher prison sentences.

Five additional independent variables were not statistically signiﬁcant (p < 0.05).
Whether an individual was on either parole or probation, whether the individual was
incarcerated at the time of the offense or had prior prison incarcerations, and whether a
PSI was conducted did not signiﬁcantly inﬂuence the amount of prison time the offender
received. With respect to the remaining independent variables several signiﬁcant
ﬁndings were observed. Older offenders were signiﬁcantly more likely to receive longer
' sentences, and so were individuals with prior jail incarcerations. Further, offenders
convicted of drug, property, and violent offenses all received more prison time than
individuals convicted of public order offenses. Convicted offenders who were out on
bond at the time of the crime were signiﬁcantly more likely to receive long prison
sentences.

Before proceeding to the discussion section one additional note is necessary.
Offenders guilty of ﬁrearm possession at the time of their present offense were
signiﬁcantly more likely to be given longer prison sentences. The regression coefﬁcient
indicates that possessing a ﬁrearm increased sentence length by thirteen years. This

ﬁnding is consistent with prior (Hagan, 1974).

32

Table 5
OLS Modelfor Sentence Length (N = 2.065)

 

 

C oefﬁcient SE t-Ratio
Constant 8.05 1 1.62 .69
Black .29 .55 .53
Hispanic 1.40 1.61 .86
Other Race -.48 2.83 -.17
Gender 1.41 1.02 1.37
Age .09 .03 3.34"
Firearm Possession 15.83 1.16 13.57"
Out On Bond -2.46 1.01 -2.43*
Parolee .33 .86 .39
Probationcr -.18 .62 -.29
Incarcerated -1.43 1.39 -1.03
Offender
Prior Prison -.96 .62 -1.54
Prior Jail -.41 .11 -3.71**
Drug Offender 3.00 .77 3.86"
Violent Offender 1.38 .69 197*
Property Offender -3.02 .85 -3.53**
PS1 -.76 1 1.54 -.06
F 19.82**
R-Square .13

 

*p<.05,**p<.01.

To summarize, two important ﬁndings were revealed in the multivariate analysis.

First, an offender’s race/ethnicity did not signiﬁcantly inﬂuence whether they were

sentenced to prison, nor did it inﬂuence the length of their sentence. Second, “legalistic”

variables were more signiﬁcant predictors of sentence disposition (i.e., whether an

offender was sentenced to prison) and sentence length. In conclusion, the multivariate

analyses indicate that an offender’s race does not inﬂuence sentencing outcomes.

33

DISCUSSION

The present analysis compared blacks, Hispanics, and members of“other”
minority groups with white offenders to answer two questions: Do blacks get sentenced
to prison more often than whites? Do blacks receive harsher sentences? The study
produced two important ﬁndings. First, African-Americans were not sentenced to prison
more often than whites, nor did they receive longer periods of incarceration. Second,
“legalistic” characteristics (e. g., offense type, prior record, legal status) were more
signiﬁcant predictors of sentence disposition (i.e., whether an offender was sentenced to
prison) and sentence length than an individual’s race. In recent years, social scientists
have debated whether legal or extra-legal characteristics are more important during
criminal sentencing procedures. The results reported here suggest that legal variables are
more important. Offense type, weapon use, legal status and prior record were found to be
signiﬁcant predictors. The results of the present analyses parallel Spohn’s (1990)
ﬁndings that sentencing decisions are determined primarily by the seriousness of the
crime committed and by the offender prior criminal record (i.e., legal variables). Other
factors that fell outside of the legalistic heading were also signiﬁcant determinants. For
example, an individual’s gender and age at the time of their present offense strongly
inﬂuenced sentence dispositions. These observations suggest that some early criticisms
of extra legal variables (age and sex) were premature (LaFree, 1985: 298-306; Hagan,
1974)

Some individuals might interpret the present analysis to suggest that sentencing
decisions are based solely on legally prescribed characteristics, but I disagree. The

differential treatment of blacks may not be seen during sentencing but discrimination may

34

be uncovered in earlier stages ofthe criminal justice process (i.e., prior to sentencing in
criminal courts). For example, scholars argue that determinant sentencing may
encourage prosecutors to exercise more discretion when deciding which cases are worthy
of full prosecution (Zatz & Hagan, 1985). Though not supported in the present analysis,
conﬂict theorists contend that prosecutors who are mainly white and representatives of
the ruling class will exercise their legal authority disproportionately on minority
defendants (Zatz & Hagan, 1985).

Peterson and Hagan’s (1984) study on sentencing in drug-related cases was
conducted during the late 19605 and 19703. As noted, during this time the Nixon
administration reformed federal drug statues. The authors contended that the American
drug prohibition began with the portrayal of African-Americans and “other” minority
groups as the “villains responsible for the growing drug menace” (Peterson & Hagan,
1984: 67). This connection between blacks and drug offenses held true in the present
analysis. It was observed that blacks were disproportionately involved in drug crimes
when compared with other types of crimes (i.e., property and violent crimes). Offenders
guilty of drug crimes were signiﬁcantly more likely to receive harsher sentences than
other types of offenders. Further, correlations were observed between African-
Americans and ﬁrearm possession. Firearm possession was the most signiﬁcant predictor
of sentencing outcomes.

It should be noted that the two ﬁndings pertaining to black offenders and legal
variables (i.e., drug offenses and ﬁrearm possession) are supported by Richard Quinney’s
(1970) six propositions of the “social reality of crime.” For the purposes here, two

propositions will be discussed. Quinney (1970: 15) posited “crime is a deﬁnition of

35

human conduct that is created by authorized agents in a politically organized society.”
Here the author contends that those with political power deﬁne or label what is criminal.
Is it outside the realm of possibility that the Nixon Administrations war on drugs targeted
“inner city” neighborhoods realizing that harsh drug laws would disproportionately affect
blacks? “Criminal deﬁnitions describe behaviors that conﬂict with the interest of
segments of society that have the power to shape public policy” (Quinney, 1970216).
Here, the author posits that those in power determine what is criminal and make laws to
address the behaviors deﬁned as such. Prior research has noted that stiffer penalties for
drug violations and the use or possession of a ﬁrearm while committing felonies have
resulted in longer prison sentences in Michigan (Early, 2001). Again, one could
speculate that African-American offenders propensity towards ﬁrearm possession, which
was supported by the present study, could be targeted by policy makers, resulting in more
stringent gun law penalties.

Why is it that race was not found to be a signiﬁcant predictor of sentencing
outcomes in the present study? One explanation is that Michigan’s judges have lost the
majority of their discretion due to highly structured sentencing guidelines, hence the
chances for discrimination based on an offender’s race is low. This process starts with
the sentencing information request (SIR) (Michigan Sentencing Guidelines, 1999). The
SIR contains all pertinent information regarding the sentencing of an offender. Each
offense covered by the guidelines is divided into separate crime groups. The groups are
as follows: Assault, Burglary, Criminal Sexual Conduct, Drug, Fraud, Homicide,
Larceny, Property Destruction, Robbery and Weapons Possession (Michigan Sentencing

Guidelines, 1999). The offender’s prior criminal record and present offenses are scored

36

using a point system. An individual’s score is then matched up against a sentence length
grid. Each grid provides a recommended minimum sentence range for combinations of
the offense and prior record levels. With each crime group there is a sentencing guideline
grid for the offender’s present crime.

A perceived ﬂaw in the guidelines is that judges are afforded broad sentencing
discretion with regards to “habitual offenders.” The established guidelines do not include
habitual offender sentencing parameters. Hence, in order to derive a sentence length for
habitual offenders, judges must “score” the offender and establish a new statutory
minimum and maximum (Michigan Sentencing Guidelines, 1999). Through the present
analysis it was observed that positive correlations exist between African-Americans and
prior criminal involvement (e. g., prior prison incarcerations). Black offenders may be
more likely than whites to have more extensive histories of criminal behavior (i.e.,
habitual offenders) and may be punished more severely (Bridges & Crutchﬁeld, 1988).
However, the results of the present analysis do not support this conclusion. It is my
contention that this rational is the most credible explanation of why an offender’s race
was not found to be a signiﬁcant predictor of sentence length or the decision to
incarcerate in Michigan for 1998.

Some additional notes are necessary at this time. Once again, Quinney’s
(1970:15) ﬁrst proposition (i.e., “crime is a deﬁnition of human conduct that is created by
authorized agents in a politically organized society”) is useful. Although judges’
sentencing practices in Michigan are highly structured, the guidelines they employ could
reﬂect racial bias. Certain offenses carry harsher penalties (i.e., drug offenses and

ﬁrearm possession)(Early, 2001 ). Through the present analysis it has been observed that

37

African-Americans disproportionately commit offenses that fall into the crime categories
mentioned above (e.g., drug offenses). Scholars contend that policy makers target these
areas (i.e., drug offenses and ﬁrearm possession) to enforce the most stringent sentencing
penalties (Early, 2001).

As previously noted, Thorston Sellin’s (1938) culture conﬂict is helpful in
conceptualizing the issue. Primary cultural conﬂict (Sellin, 1938) between African —
Americans and policy makers is a plausible explanation for the racial discrimination
argument. Previous research has noted that the conduct norms of African—Americans
versus policy makers (mainly white) conﬂict, hence resulting in the formulation of
criminal law that targets certain behaviors (Sellin, 1928). Further, such behaviors (i.e.,
drug offenses and ﬁrearm possession) could be speciﬁcally targeted as external
manifestations of racial discrimination (by policymakers) and carry more harsh
sentencing outcomes (Early, 2001).

Although the present analysis did not ﬁnd race to inﬂuence sentencing outcomes,
I believe it would be premature to posit that discrimination in the criminal justice system
is no longer present. Instead these results should encourage other individuals to also
question why race was not found to factor into sentencing decisions in Michigan.
Ultimately, racial discrimination may no longer be overtly present in the highly
formalized and publicized environment of the American courtroom. However,
discrimination may still persist elsewhere in the criminal justice system, and should be

revealed. targeted, and eradicated.

38

BIBLIOGRAPHY

Albonetti, CA. 1991. Integration oftheories to explainjudicial discretion. Social
Problems 38:247-262.

Bachman, R. and R. Paternoster. 1997. Statistical Methods for C riminologi' and
Criminal Justice. New York: McGraw-Hill.

Becker, HS. 1963. Outsiders: Studies in the Sociology of Deviance. New York: Free
Press.

Bernstein, I., W. Kelly, and P. Doyle. 1977. Societal reaction to deviants: The case of
criminal defendants. American Sociological Review 42:743-755.

Blumstein. A. 1982. On the racial disproportionality of United States’ prison
populations. The Journal of Criminal Law & Criminology 73:1259-1281.

Blumstein, A., J. Cohen. S. E. Martin, and M.H. Tonry. 1983. Research on Sentencing:
The Search for Reform. Vol. 1. Washington, DC: National Academy Press.

Bridges, GS. and RD. Crutchﬁeld. 1988. Law, social standing and racial disparities in
imprisonment. Social Forces 66:699-722.

Bullock, HA. 1961. Signiﬁcance of the racial factor in the length of prison sentences.
‘ Journal of Criminal Law, Criminologv & Police Science 52:41 1-417.

Bureau of Justice Statistics. 1998. Prison and Jail Inmates at Midyear 1998.
Washington, DC: US. Department of Justice.

Burke, RI. and AT. Turk. 1974. Factors affecting post-arrest dispositions: A model for
analysis. Social Problems 22:313-330.

Chambliss, W.J. and J.T. Liell. 1966. The legal process in the community setting: A
study of law enforcement. Crime & Delinquency 12:310-317.

Chambliss, W. and RT. Seidman. 1971. Law, Order and Power. Reading, MA:
Addison-Wesley.

Chiricos, T.G., P.D. Jackson, and GP. Waldo. 1972. Inequality in the imposition of a
criminal label. Social Problems 19:553-572.

Early, K. 2001. Criminal Justice in Michigan. Needham Heights, MA: Allyn & Bacon.

Farrell, RA. and V.L. Swigert. 1978. Legal disposition of inter-group and intra-group
homicides. The Sociological Quarterly 19:565-576.

39

Gibson, J.L. 1978. Race as a determinant of criminal sentences: A methodological
critique and a case study. Law & Societv Review 12:455-478.

Goffman, E. 1963. Stigma: Notes on the Managenrent of Spoiled Identity. Englewood
Cliffs, NJ: Prentice-Hall

Green, E. 1964. Inter- and intra-racial crime relative to sentencing. Journal of Criminal
Law, Criminologv & Police Science 55:348-358.

Griswold, D. 1987. Deviation from sentencing guidelines: The issue of unwarranted
disparity. Journal ofCriminal Justice 15:317-329.

Hagan, J. 1974. Extra-legal attributes and criminal sentencing: An assessment of a
sociological viewpoint. Law & Society Review 81357-383.

Hagan, J. 1979. Criminal Justice in rural and urban communities: A study of the
bureaucratization of justice. In S.L. Messinger and E. Bittner (eds.), Criminology Review
Yearbook. Beverly Hills, CA: Sage.

Hagan, J. and C. Albonetti. 1982. Race, class and the perception of criminal injustice in
America. American Journal of Sociology 88:329-355.

Hawkins, D. 1987. Beyond anomalies: Rethinking the conﬂict perspective on race and
criminal punishment. Social Forces 65:719-745.

Hindelang, M.J. 1969. Equality under the law. Journal of Criminal Law, Criminology
& Police Science 60:306-313.

Kleck, G. 1981. Racial discrimination in criminal sentencing: A critical evaluation of
the evidence with additional evidence on the death penalty. American Sociological

Review 46:783-805.

Kleck, G. 1985. Life support for ailing hypothesis: Modes of summarizing the evidence
for racial discrimination in sentencing. Law & Human Behavior 3:271 -285.

Kramer, J .H. and J .T. Ulmer. 1996. Sentencing disparity and departures from
guidelines. Justice Quarterly 13281-103.

LaFree, G. 1985. Adversarial and nonadversarial justice: A comparison of guilty pleas
and trials. Criminology 23:289-312.

Levin, MA. 1977. Urban Politics and The Criminal Courts. Chicago and London: The
University of Chicago Press.

40

Lilly, R.J., F.T. Cullen. and RA. Ball. 1989. Crinrinological Theory Context and
Consequences. Newbury, CA: Sage.

Lizotte. A]. 1978. Extra-legal factors in Chicago’s criminal courts: Testing the conflict
model ofcriminal justice. Social Problems 25:564-578.

Martindale, D. 1960. The Nature and Types ofSociological Theory. Cambridge, MA:
Houghton Mifﬂin.

Marx, K. and F. Engels. 1848. Marx-Engels Selected Works, Vol.1. London: Lawrence
and Wishart.

Michigan Sentencing Guidelines Manual. 1999. St. Paul, MN: West Group.

Miethe, TD. and CA. Moore. 1986. Racial differences in criminal processing: The
consequences of model selection on conclusions about differential treatment. The
Sociological Quarterly 27:217-237.

Norusis, M.J. 2000. SPSS 10.0: Guide to Data Analysis. Upper Saddle River, NJ:
Prentice—Hall.

Peterson, R. and J. Hagan. 1984. Changing conceptions of race: Toward an account of
anomalous ﬁndings of sentencing research. American Sociological Review 49:56-70.

Platt, AM. 1969. The Child Savers: The Invention ofDelinquency. Chicago:
University of Chicago Press.

Pope, C .E. 1975. Sentencing of California Felony Offenders, Analytic Report #6.
Albany, NY: Criminal Justice Research Center.

Pratt, TC. 1998. Race and sentencing: A meta-analysis of conﬂicting empirical
research results. Journal of Criminal Justice 26:513-523.

Quinney, R. 1970. The Social Reality ofCrime. Boston, MA: Little Brown.

Reiman, J. 1979. The Rich Get Richer and the Poor Get Prison. New York: John
Wiley.

Schur, EM. 1965. Crimes without Victims: Deviant Behavior and Public Policy.
Englewood Cliffs, NJ: Prentice—Hall.

Sellin, T. 1928. The Negro Criminal: A Statistical Note. The American Negro.
Philadelphia, PA: The American Academy of Political and Social Science.

Sellin, T. 1938. Culture Conﬂict and Crime. New York: Social Science Research
Council.

41

Smith, J. and W. Fried. 1974. The Uses of The American Prison: Political T heory and
Penal Practice. Lexington, MA: Lexington.

Spohn, C. 1990. The sentencing decisions of black and white judges: Some expected
and unexpected similarities. Law & Society Review 24:] 197-1216.

Spohn, C., J. Gruhl, and S. Welch. 1981. The effect of race on sentencing: A
reexamination of an unsettled question. Law & Society Review 16:71-88.

Stone, G. 1962. Appearances and the self In A. Rose (ed.), Human Behavior and
Social Processes. Boston, MA: Houghton Mifﬂin.

Swigert, V.L. and RA. Farrell. 1976. Murder, Inequality and the Law: Differential
Treatment in the Legal Process. Lexington, MA: D.C. Heath.

Swigert, V.L. and RA. Farrell. 1977. Normal homicides and the law. American
Sociological Review 42:16-32.

Thomas, CW. and J .R. Hepburn. 1983. Crime, Criminal Law, and Criminology.
Dubuque, IA: W.C. Brown.

Turk, AT. 1969. Criminalin and Legal Order. Chicago: Rand McNally

Turk, AT. 1969. Introduction. In W. Bonger, Criminalin and Economic Conditions.
Bloomington, IN: Indiana University Press.

Uhlman, TM. and ND. Walker. 1979. A plea is no bargain: The impact of case
dispositions on sentencing. Social Science Quarterb’ 60:219-232.

Unnever, J.D., C.E. Frazier, and J.C. Henretta. 1980. Race differences in criminal
sentencing. The Sociological Quarterly 21 :197-205.

Vold, GB. 1958. Theoretical Criminology. New York: Ox ford University Press.

Vold, GB. and T.J. Bernard. 1986. Theoretical Criminology. 3rd ed. New York:
Oxford University Press.

Walker, 8., C. Spohn, and M. DeLone. 1996. The Color Of Justice: Race, Ethnicity,
and Crime in America. Belmont, CA: Wadsworth.

Walker, 8., C. Spohn, and M. DeLone. 2000. The Color Of Justice: Race, Ethnicity,
and Crime in America, 2nd ed. Belmont, CA: Wadsworth.

Wilbanks, W. 1987. The Myth of a Racist Criminal Justice System. Monterey, CA:
Brooks/Cole.

42

Wilson, W .J . 1987. The Truly Disadvantaged: The Inner City, the Underclass, and
Public Policy. Chicago: University of Chicago Press.

Wilson, J .Q. and R. J. Hemstein. 1985. Crime and Human Nature. New York: Simon
and Schuster.

Zatz, MS. and J. Hagan. 1985. Crime, time, and punishment: An exploration of
selection bias in sentencing research. Journal of Quantitative Criminology 1:103-122.

 

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

1111111311111111111217111

O 427

11