Citation
Are drug policies black and white?

Material Information

Title:
Are drug policies black and white? racial differences, stratification, and the distribution of the law
Creator:
Collins, Joy Charlotte
Publication Date:
Language:
English
Physical Description:
viii, 60 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Drugs of abuse -- Law and legislation -- United States ( lcsh )
Drug control -- United States ( lcsh )
Drugs -- Law and legislation -- United States ( lcsh )
Crime and race -- United States ( lcsh )
Discrimination in criminal justice administration -- United States ( lcsh )
Crime and race ( fast )
Discrimination in criminal justice administration ( fast )
Drug control ( fast )
Drugs -- Law and legislation ( fast )
Drugs of abuse -- Law and legislation ( fast )
United States ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 56-60).
General Note:
Department of Humanities and Social Sciences
Statement of Responsibility:
by Joy Charlotte Collins.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
785825543 ( OCLC )
ocn785825543
Classification:
LD1193.L65 2011M C65 ( lcc )

Full Text
ARE DRUG POLICIES BLACK AND WHITE?
RACIAL DIFFERENCES, STRATIFICATION, AND THE DISTRIBUTION OF THE LAW
by
JOY CHARLOTTE COLLINS
B.A Metropolitan State College of Denver 2006
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Masters of Social Science
2011


This thesis for the Masters of Social Science
degree by
Joy Charlotte Collins
has been approved by:
Stacey Bosick PhD., University of Colorado Denver,
College of Liberal Arts and Sciences
Colorado Denver,
College of Liberal Arts and Sciences
*<1 > t' it- r\ V "" -N- \
Date


Collins, Joy Charlotte MSS
Are Drug Policies Black and White?
Racial Differences, Stratification, and the Distribution of Law
Thesis directed by MG Director, Mary Dodge
ABSTRACT
This thesis uses the language of Black's theory of the Behavior of Law to apply a
quantifiable view at the relationship between race and drug crimes. Using secondary
data from the State Court Processing Statistics, 1990-2006 and the vertical dimension of
stratification; this research explores the differences between White and Black offenders,
their ability to obtain adequate council, how being released or detained affects their
penalty outcome and how these variables are related. According to Black (1976),
sociological factors such as socioeconomic status strongly influence how the law
(government social control) behaves. There are two ways the law can vary: in amount
and in style. This thesis explores both of these ways by measuring predictive value of
specific socioeconomic variables using logistic regression models and using correlations
to show how the variables interact with each other.
This abstract accurately represents the content of the candidate's thesis. I recommend
its publication.
Approved:


TABLE OF CONTENTS
Figures.......................................................................vi
Tables........................................................................vii
CHAPTER
1. INTRODUCTION..........................................................1
Research Questions..............................................2
2. REVIEW OF THE LITERATURE..............................................3
Discrimination From the Lens of the Individual..................4
Law and Judicial Discretion.....................................8
Policy.........................................................14
3. THE ROLE OF THEORY............................................17
4. METHODOLOGY..........................................................20
Statistical Methods............................................21
Indicator Variables............................................24
Characteristics of the Populations.............................24
5. MINOR DRUG CRIMES....................................................29
Correlations...................................................29
Minor Drug Crimes 2002.........................................32
Minor Drug Crimes 2004.........................................34
IV


Minor Drug Crimes 2006
37
6. MAJOR DRUG CRIMES...................................................40
Correlations..................................................40
Major Drug Crimes 2002........................................43
Major Drug Crimes 2004........................................45
Major Drug Crimes 2006........................................47
7. CONCLUSIONS.........................................................50
Limitations...................................................54
APPENDIX A
A. Operating definitions...............................................53
REFERENCES..................................................................54
DATA REFERENCES.............................................................58
v


LIST OF FIGURES
Figure 1. Variable Model...........................................................23
Figure 2. Variable Coding..........................................................24
VI


LIST OF TABLES
Table 1. Minor drug crimes by race 2002-2006.....................................25
Table 2. Major drug crimes by race 2002-2006.....................................25
Table 3. Legal representation by race 2002-2006-minor drug crimes................26
Table 4. Legal representation by race 2002-2006- major-drug crimes...............26
Table 5. Release and detention status by legal representation 2002-2006-
Minor drug crimes.....................................................27
Table 6. Release and detention status by legal representation 2002-2006-
Major drug crimes.......................................................28
Table 7. Variable by variable correlations-Minor drug crimes...................30
Table 8. Variable correlations of independent variables to penalty received for minor
drug crimes 2002-2006...................................................31
Table 9. Predicted and observed values Minor drug crimes 2002..................33
Table 10. Logistic regression predicting sentencing outcomes-
Minor drug crimes 2002...............................................34
Table 11. Test of Model Coefficients-Minor drug crimes 2002....................34
Table 12. Predicted and observed values Minor drug crimes 2004.................36
Table 13. Logistic regression predicting sentencing outcomes-
Minor drug crimes 2004.................................................36
vii


Table 14. Test of Model Coefficients-Minor drug crimes 2004........................37
Table 15. Predicted and observed values Minor drug crimes 2006.....................38
Table 16. Logistic regression predicting sentencing outcomes-
Minor drug crimes 2006...................................................38
Table 17. Test of Model Coefficients-Minor drug crimes 2006.......................38
Table 18. Variable by variable correlations-Minor drug crimes.....................41
Table 19. Variable correlations of independent variables to penalty received for
Major Drug crimes; 2002-2006.............................................42
Table 20. Predicted and observed values Major drug crimes 2002.....................44
Table 21. Logistic regression-Major drug crimes 2002...............................44
Table 22. Tests of Model Coefficients- Major drug crimes 2002......................45
Table 23. Predicted and observed values-Major drug crimes 2004.....................46
Table 24. Logistic regression predicting sentencing outcomes-
Major drug crimes 2004...................................................46
Table 25. Test of Model Coefficients-Major drug crimes 2004........................47
Table 26. Predicted and observed values-Major drug crimes 2006.....................48
Table 27. Logistic regression predicting sentencing outcomes-
Major drug crimes2006....................................................48
Table 28. Test of Model Coefficients-Major drug crimes 2006........................49
viii


CHAPTER I
INTRODUCTION
According to Donald Black (1976, 2010) "Law," in the language of pure sociology,
"is governmental social control. It is, in other words, the normative life of a State and its
citizens, such as legislation, litigation, and adjudication" (p. 2). Law is a quantitative
variable that can be measured empirically and, "the quantity of law is known by the
number and scope of prohibitions, obligations, and other standards to which people are
subject, and by the rate of legislation, litigation, and adjudication" (p. 3). As more law is
applied to an individual, that individual's ability to live their day to life decreases.
Donald Black's theory of the Behavior of Law posits that the law behaves
differently based on different sociological factors. Alexander (2010), for example,
suggests that unfair enforcement and unfair sentencing in drug cases has created a
modified version of the Jim Crow laws. According to Moore and Elkavich (2008), in 1968
"The war on drugs," designed to redouble efforts against the sale, distribution, and
consumption of illicit drugs in the United States was instituted by the Nixon
administration and has been embraced by subsequent administrations, has created a
large mechanism of courts, jails, and prisons and failed to decrease the use of drugs
while doing much to create confusion and hardship in families of color and urban
communities.
1


Stratification including poverty, location, education, and racial make-up plays a
primary role in Donald Black's theory and also has an effect on what type of defense one
is able to mount on one's own behalf; this is true even when dealing with the most
minor of charges. The difference between being released or detained before going
before a judge at the pretrial hearing can affect your sentencing outcome and your legal
representation may affect whether you are detained or released before going before
the judge; this can include bail or being released on your own recognizes. The penalty
outcomes related to legal representation and being released or detained ultimately
circle back to stratification and how the law behaves differently given these vertical
layers of social life that are primarily determined by wealth (Black, 1976). While it is
impossible to measure the intentions behind drug policies and if they are really meant
to limit African American Civil Rights at a greater rate than those of White offenders, the
application of Black's theory to Alexander's concept of the reemergence of Jim Crow
laws indicates there is a need for policy change. This thesis addresses two research
questions:
1. Does Black's theory of the Behavior of Law as it pertains to stratification explain
any differences in the distribution of drug offense penalties between Black and
White offenders?
2. Does legal representation mediate the relationship between race and penalty
outcomes in major and minor drug offense cases and/or does being detained or
released prior to going before a judge mediate ones penalty outcome?
2


CHAPTER II
REVIEW OF THE LITERATURE
Beckett, Nyrop, and Pfingst (2005), conducted a study using multiple data sources
including the Seattle Needle Exchange Survey to explore race/ethnicity, the drug(s)
present in the needle(s) just exchanged, and the race/ethnicity of the person from
whom they had obtained those drugs. The research included data from treatment
centers, drug related arrests, and Arrestee Drug Abuse Monitoring System (ADAM) data.
The Seattle Needle Exchange Survey produced 589 surveys providing information on
over 800 drug transactions. According to Beckett, Nyrop, and Pfingst (2005), most
researchers examining whether or not arrests represent an accurate measure of
unlawful behavior have concluded that race plays a comparatively small role in arrests
for serious offenses, such as murder and robbery, but a larger role in the policing of
minor offenses. Moreover, when the parties directly involved in the illicit behavior are
consenting, more proactive and discretionary law enforcement techniques are used.
These findings imply that drug enforcement may be shaped by race. Comparative data
show that the focus on crack cocaine contributed to overrepresentation of minorities
(Beckett, Nyrop, & Pfingst, 2005). This has been an argument that gathered support
with the advent of "the war on drugs" and has been a frame of many research studies;
who is using drugs and who is being punished for using drugs?
3


Discrimination from the Lens of the Individual
Analyzing data from the 1987 National Opinion Research Center (NORC) General
Social Survey, Cohn, Barkan, and Halteman, (1991) found that the punitive attitudes of
Whites towards criminals are based partly on racial prejudice, while those of Blacks
stem from their fear of crime. The bivariate analysis found that Blacks were 9.6% less
likely than Whites to think that the courts are too harsh. These findings are inconsistent
with many conflict theories that argue that Blacks and poor people tend to be less
punitive and engage in more crime than White non-poor people (Cohn, Barkan, &
Halteman, 1991).
Using data from the National Race and Crime Survey administered by the Survey
Research Center, Hurwitz and Peffley (2005) explored the differences in citizens' beliefs
about the fairness of the United States justice system. In order to investigate this,
respondents were asked to indicate their level of agreement with two statements: "The
justice system in this country treats people fairly and equally" and "The courts in your
area can be trusted to give everyone a fair trial." Responses to both items showed vast
differences in beliefs between races with an average of 67.5% of African American
respondents and 32.5% of White respondents disagreeing with the two statements.
While 74% of Blacks disagree that the justice system treats everyone equally, only 44.3%
of Whites feel the same. The racial differences are more acute when observing
responses to the second question "the courts in your area can be trusted to give
everyone a fair trial," with 61% of Black and 26% of Whites who do not trust the courts
4


to give a fair trial. These different beliefs of fairness have implications on how
individuals behave in different situations. Using these assumptions, Hurwitz and Peffley
(2010) conducted the "Police Brutality Experiment." Respondents heard about an
incident involving confrontations between police and civilians that manipulated the race
of the civilian. The assumption is respondents are more sympathetic to targets of their
own race. Blacks who thought that the justice system was unfair were more than twice
as likely to think that the police investigation would be fair when the motorist was
White (55%) than if he was Black (24%). This finding does not mean that Blacks
habitually favored their race but rather they tended to side with the Black civilian if they
believed that the criminal justice system was chronically unfair. Whites were found to
punish officers equally regardless of the race of the victim. When faced with a racially
charged case, they were even more likely to punish the officer. Sigelman and Welch
(1991) introduced a concept that Blacks and Whites inhabit different perceptual worlds
and while the Black respondents expected the unfair behavior, Whites were surprised
by it and responded by punishing harshly. Sigelman and Welch argued that the (correct)
belief that the criminal justice system is inherently biased against the African American
community encourages suspicious interpretations of all police behavior and influences
all interactions with law enforcement (Hurwitz & Peffley 2005).
Using data from the Washington Post, Henry J. Kaiser Family Foundation and
Harvard University 2006 African American Survey, Unnever (2008) focused on three
related issues: (a)whether African Americans and Whites share a common "sensibility"
5


or "cognitive landscape" when considering why African-American men are
disproportionately imprisoned, (b) whether the opinions held by African Americans and
Whites are collectively held, and (c) whether the relative subordinate position of African
Americansas created by their personal experiences with racial discrimination shapes
the opinions that they have about why Black men are disproportionately incarcerated.
Unnever based this research on the comparative conflict theory posited by Hagan,
Shedd, and Payne (2005) states: "scholars need to be sensitive to and need to
incorporate into their models of public opinion the "nuanced understanding of the
relative positions of subordinate and dominate groups who form relative and collective
points of reference for one another (p. 513)." Unnever argued that a defining aspect of
the subordinate position African Americans have in relation to Whites is their
experience with racial discrimination.
The results indicated that 71% of the African American respondents were more
likely to believe that the "big reason" Blacks are incarcerated is police bias, whereas 63%
of Whites were most likely to believe it results from poverty. The "big reason" results
also indicated that African Americans are least likely to believe that the disproportionate
imprisonment of Black men results from ignorance of the law on the part of African
American offenders (42%). In addition, 28% of Whites are less likely to believe that Black
male incarceration is related to the courts convicting more African Americans than
Whites and can more likely be attributed to Black men not knowing that crime is wrong
(28%). Interestingly, the results indicated that African Americans are more likely than
6


Whites to report that Black male imprisonment is related to African-American parents
failing to teach their children right from wrong and from Black men believing crime is
not wrong. Overall, the study indicated that that the cognitive landscape that African
Americans collectively hold about why Black men are incarcerated is likely shaped by
their personal experiences with racial discrimination (Unnever, 2008).
Lee and Rasinski (2006) conducted a study using a national probability sample of
White Americans to examine the effects of moralism, attributions of blame, beliefs
about racial group use of cocaine, beliefs about racism, and beliefs about the
effectiveness of law enforcement on judgments about the appropriate sanction for
being caught for the first time with five grams of cocaine. The predicted direct
relationship between racism and severity of sanction was not found; however, a strong
indirect relationship was discovered as racism operated through morality, blame, and
beliefs about racial group use of cocaine. The indirect relationship indicated that racial
sentiments influence moral sentiments and moral sentiments influence support of
punitiveness. When asked about sanctions for first time offenders caught with five
grams cocaine, 51% of the sample indicated that treatment would be an appropriate
sentence; however those who believed Blacks were the primary consumers also favored
more punitiveness (Lee & Rasinski, 2006).
Conflict theory and previous research suggest that the Black-White difference in
support for harsh criminal punishments may be linked to anti-Black prejudice among
Whites and perceived injustice among Blacks. Using survey data from the 2001 Race,
7


Crime and Public Opinion Study, Johnson (2008) examined the sources of the racial gap
in levels of punitiveness. Johnson tests two main hypothesis; perceived racial bias in the
criminal justice system and racial prejudice. Using a representative sample of 978 non-
Hispanic White and 1,010 non-Hispanic Black respondents living in U.S. households,
Johnson built on a large body of work linking racial bias and perceived racial bias in
disparities in punitive attitude among Blacks and Whites. Overall, Whites scored higher
than Blacks on four predictors of punitiveness: (a) favor trying juveniles as adults, (b)
parole boards should be stricter, (c) favor three strike laws, and (d) penalties for violent
crimes are too lenient. The results showed that perceived racial bias significantly
reduced support for punitive policies, while negative racial stereotyping and racial
resentment had a significant positive effect on support for harsh penalties. These results
indicated that, together, racial prejudice and perceived racial bias explained the Black-
White gap in punitive attitudes (Johnson 2008).
Law Enforcement and Judicial Discretion
Chappell and Maggard (2007) used data from the New York City Courts and
Black's Theory of Law to help understand discrepancies in crack and cocaine case
dispositions. Their findings revealed that racial minorities are more likely to be charged
with a felony and receive longer prison sentences compared to Whites while showing
only mixed support for Black's arguments. This study used quantity of law as a
dependent variable and multiple independent variables to represent stratification,
respectability, and conventionality. Overall, Chappell and Maggard found support for
8


Black's measures of conventionality and respectability. In addition, the results showed
major racial discrepancies in the dispositions of cases in the sample.
David Mustard (2001) examined 77,236 federal offenders sentenced under the
Sentencing Reform Act of 1984 and after controlling for extensive criminological,
demographic, and socioeconomic variables found that Blacks, black males, and
offenders with low levels of education and income receive substantially longer
sentences. These disparities are generated by departures from the standard sentencing
guidelines, rather than differential sentencing within the guidelines. Mustard found
that Blacks and Hispanic received longer sentences than Whites (5.5, 4.5, and 2.3
months respectively). In addition, females received 5.5 fewer months than males. The
average sentence length was 46 months, so evaluated at the mean, Blacks receive about
12% longer terms than Whites, and males receive 12% longer terms than females. When
these data were examined by offense type, the racial disparities were more pronounced
for certain offenses. The percentage difference was greatest for those convicted of drug
trafficking, where Blacks are assigned sentences 13.7% longer than Whites. The study's
primary conclusion was that after including more exhaustive controls than any previous
study, large differences in the length of sentences exist on the basis of race, gender,
education, income, and citizenship; this despite explicit statements in the Federal
Sentencing Guidelines that indicate that these characteristics should not affect any
sentence length. In addition, Blacks and males are less likely to get no prison term when
that option is available; less likely to receive downward departures; more likely to
9


receive upward adjustments and, conditioned on having a downward departure, receive
smaller reductions than Whites and females (Mustard 2001).
As of 2006, 56% of federal prison inmates were incarcerated for narcotics
offenses. Research conducted by Hartley (2008) analyzed the decision-making practices
of judges for narcotic violations in four districts in the southwestern United States.
Three judicial decisions were analyzed: the length of the prison sentence imposed on
defendants sentenced to prison, the magnitude of the sentence discount or lowering
from the guideline sentence that an offender received for a regular downward
departure (from the sentencing guidelines), and the magnitude of the sentence discount
that an offender received for a substantial assistance departure (assisting in prosecution
of others). This study used data from the Interuniversity Consortium for Social and
Political Research and included defendants sentenced in 2003 for narcotics trafficking
offenses from four states: Texas, New Mexico, Arizona, and California. The length of
sentences increase as the seriousness of the crime increases, however, the results
showed that sentencing varied across jurisdictions due to judicial discretion. In the state
of California, when a crime fell under minimum sentencing, offenders faced greater
sentences whereas in Texas this effect was opposite. These results were unexpected;
the presence of a mandatory minimum sentence should have had a significant and
positive effect on sentences for narcotics offenders across districts. The only area that
race and ethnicity affected judicial discretion involved sentences that included
substantial assistance departures. In these cases, Hispanics received sentencing
10


discounts that were 8.5% less than those received by Caucasians. Overall, it was found
that race, ethnicity, gender, and citizenship status, in some contexts, either directly or
indirectly through departures, continued to be significant predictors of outcomes in the
federal criminal justice system (Hartley, 2008).
Radosh (2008) reviewed trends in current incarceration patterns in light of the
disproportionate rise in incarcerations for drug crimes. These high rates of incarceration
are partially due to several acts passed as part of "the war on crime" and then later "the
war on drugs." The minimum mandatory penalties in the 1984 Sentencing Reform Act,
the 1984 Comprehensive Crime Control Act, the 1986 Anti-Drug Abuse Act, and the
1988 Anti-Drug Abuse Act have set the penalties for crimes typically committed by
African American or Hispanic offenders significantly longer than crimes typically
committed by White offenders. The penalty for selling 50 grams (1.76 ounces) of crack
cocaine is 10 years to life, while the penalty for selling 5 kilograms (10.6 pounds) of
White powder cocaine is 10 years to life. African American or Hispanic offenders
commonly sell crack, while the sale of powder cocaine is a typically committed by
Whites. In the United States, 3218 African American men out of every 100,000
population were incarcerated in a state or federal prison in 2004. This rate is 6.95 times
higher than the rate for White men. The trends reviewed in this article reveal a problem
that in an effort to control a crime problem, the U.S. has created a civil rights problem
(Radosh, 2008).
11


Brennan and Spohn (2008) examined sentencing outcomes from a random
sample of felony drug offenders convicted during calendar year 2000 in a large urban
jurisdiction in North Carolina. The analysis focused on Black-White, Hispanic-White, and
Hispanic-Black differences. The data showed that White offenders received less severe
punishments than either Blacks or Hispanics; Hispanic offenders were particularly
disadvantaged because they received harsher punishments relative to both Blacks and
Whites. Whites received more lenient sentences or community punishment (lowest
level) rather than the more punitive option of incarceration that Blacks and Hispanics
were likely to get (Brennan & Spohn, 2008).
To determine whether or not age and gender influence the effects of race on
prosecutorial charging decisions, Franklin (2010) used a sample of 4,349 felony drug
defendants for which information was provided by the 1998 State Court Processing
Statistics to test a theory that suggests that while prosecutors may not subscribe to the
concept of criminal stereotypes, the prosecutor trying a case might consider the
possibility that a jury will be influenced by prevalent stereotypes of the typical criminal.
To the degree that this occurs, it is possible that prosecutors will assess young Black
males as more convictable than other similar non-Black defendants. Using age as a
mediating variable, the model showed that both males and females regardless of race
were treated equally. Ultimately, Franklin found that race was conditioned by age but
not by gender. Whites between the ages of 30 and 39 were significantly less likely to
have their case dismissed than a Black defendant in that same age range regardless of
12


gender. In contrast, Barnes and Kingsnorth (1996) argued that Black defendants were
more likely to have their cases dismissed as compared to White defendants. Black
defendants were likely arrested with poorer quality evidence when compared with
White defendants, forcing prosecutors to dismiss their cases more frequently. This
study focused on felony drug cases only and showed that race did play a role in
prosecutorial dismissals. The findings were not in the direction one may expect; middle-
aged White defendants were treated more punitively than young Black defendants
(Franklin, 2010).
Schlesinger (2011) examined the effects of mandatory terms and sentencing
enhancements on Black and White men's state-level prison admission rates. The data
were constructed by merging information collected on sentencing policies through
archival research at state law libraries from several publicly available databases,
including the National Corrections Reporting Program, 1983-2000 (NCRP) and cross-
sectional time-series analysis. The research produced three major findings: (1) both
mandatory terms and sentencing enhancements increase prison admission rates for
Black and White men, (2) these policies disproportionately increase Black men's
admissions, and (3) the effects of these policies on both scale and disparity are strongest
and most consistent on admissions for violent offenses. According to Schlesinger (2011),
the findings are consistent with theories of modern racism, which argue that, in the
post-civil rights era, racial disparities are primarily produced and maintained by
colorblind policies and practices.
13


Policy
Studies have shown that being you are detained or released prior to trial has an
effect on the legal outcome in the criminal justice system. Some have argued that this is
an even stronger predictor than race (Phillips, 2008). A Canadian study done by Moyer
and Basic (2004) indicated that detention at the initial bail hearing disadvantages the
individual both in terms of the likelihood of being convicted (as a result of a guilty plea)
and of receiving a jail sentence. Moyer and Basics' regression analysis found that the
odds of pleading guilty were much higher for those who were detained than those who
were released, and that detained individuals were less likely to have all of their charges
dropped. Phillips (2008), using data for non-felony cases from the New York City
Criminal Justice Agency (CJA) from 2003-2004 (n=28,766), also found that the likelihood
of conviction appeared to be affected by detention at arraignment as well as the length
of that detention. The research found that as the detention length became longer (up to
60 days) the conviction rate increased (Phillips, 2007). Detention was found to have a
stronger influence on the likelihood of conviction if the defendant was Black or Hispanic.
Bobo (2006) argued that the United States has enacted policy changes that have
created an extraordinary rise in the use of incarceration for purposes of social control
and that these actions have had sharply disproportionate effects on African Americans.
In addition, African Americans directly experience and are aware of these changes.
Because of this are by overwhelming margins deeply disillusioned with the criminal
justice system. Many see this situation of inequality before the law as a setback of
14


progress in civil rights movement. Finally, Bobo argued that this disillusionment is
contributing to a crisis of legitimacy, a crisis that will have effects on how Blacks engage
legal authority in terms of interactions with police and with the court system. Using the
2001 and 2002 Race, Crime, and Public Opinion surveys (RCPO), as well as four focus
group discussions, the study produced strong evidence of distrust and disillusionment
with the institutions of the crime response complex and found that 89% of African
American affirmed the idea that the criminal justice system is biased against Blacks
compared to 38% of Whites (Bobo, 2006).
The literature reviewed thus far has shown consistent instances of discriminate
practices as a result of the war on drugs. This thesis builds on the current body of
literature using longitudinal nationally representative data to explore the relationship
between race, legal representation, and whether one is released or detained before
they go before the judge. In addition, by separating drug crimes by severity and
analyzing them independently using a quantitative method to measure the amount of
law applied for each type of crime based on race, legal representation, and release or
detention status at the time an individual goes before a judge allows the effect
stratification has on sentencing outcomes become clearer. The differences between
major and minor drug crimes makes a difference in resources, sentences and judicial
discretion making separate, quantitative analysis a more accurate form of analysis for
this research. This research is important because if the lowest status persons are the
most likely to be socially controlled by the government (which by definition lowers their
15


status and their control of their own environment), and, in turn, are subjected to
excessive amounts of law that reduces people's ability to gain employment, earn
money, get married, and obtain an education, then a logical conclusion is that low status
begets low status, which is in part a function of the discriminatory effect of law. In
short, social inequalities lead to the reproduction of social inequalities (Jacques &
Wright, 2010). Using Black's theory and the language of The Behavior of Law, this thesis
uses a quantified approach to viewing the relationships between race, legal
representation, and whether being released or detained has an effect on sentencing
outcomes.
16


CHAPTER III
THE ROLE OF THEORY
Donald Black (2010) posits a theory where the features of society are predictors,
and the behavior of law is the outcome that sociological factors affect how the law
behaves. He argues that the direction of law is always the opposite of the direction of
the crime. A crime is committed by a person with a lower socio-economic status (SES)
against a person with a higher SES would be considered an upward" crime and the law
applied against the offender would be "downward" (p.21). This research uses this
theory to look at "upward" crime and "downward" law related to drug charges, the
penalties associated with those charges, and the way the law behaves with White
offenders and Black offenders. Race plays a major factor in one's socio-economic status
and Black's theory would suggest that if a White upper class college student was caught
selling drugs in an upper class White area it would be considered "downward" crime
because the offender was of a higher SES that the typical resident of that area. In this
situation, the law would work upward or in the opposite direction of the crime. If this
same crime were committed in the same area by an African American male with a lower
SES, it would be seen as an "upward" crime because the offender is of a lower SES than
the typical resident and the corresponding law would be "downward."
17


This scenario is applicable to many situations and is central to Black's theory because
"downward law is always greater than upward law." Downward law is "greater" in the
sense that it is more likely to be applied and if applied, the penalty will be harsher.
By addressing the quantity and styles of law, Black attempts to explain the
variations of each. Both of these propositions form a relationship between law and
other aspects of social life such as stratification, morphology, culture, organization, or
social control. This relationship can be considered an inverse one: when the quantity of
law increases the quantity of social control of these other kinds decreases, and vice
versa. Social control defines what in a society is considered deviant behavior, and the
more social control there is the more deviant the conduct is. According to Black, the
theory of law predicts the definitions of law as well as the crime rate. For example, a
juvenile without social control at home is more likely to become criminal since crime is
defined by law and the amount and severity of law increases as other social control
decreases (Black, 2010). Stratification, the first dimension of this theory is considered to
be the vertical aspect of social life (Black, 2010). This means that society is stratified into
vertical layers mostly determined by wealth. This wealth often determines legal
advantage for one individual versus another and therefore, law is applied in different
amounts and different ways, varying directly with ones level of stratification. The more
stratification in a society the more the law there is. Deviant behavior, according to this
theory, varies inversely with rank; the lower the rank, the more deviant behavior.
Within each vertical layer, there is a hierarchical structure, which differentiates each
18


person in the layer from the others; there is always someone above you. Law is greater
when a society is more independent, and less when more intimate. Independent
societies have fewer controls as anonymity breeds independence and committing crime
anonymously is far easier if the person is a stranger. Relational distance determines this
intimacy or independence. It also predicts and explains the style of law, whether it is
punitive or non-punitive. Complete strangers are more likely to see each other as
adversaries, whereas intimates are more likely to offer help. Those individuals, who are
considered marginal, or poorly integrated in their layer, will most likely be singled out
and blamed for deviant acts (Black, 2010).
Many theorists focus on inequality and inequality in the justice system however
Donald Black's unique view of law as dependent on other variables allows for a strictly
quantifiable analysis of its behavior. By measuring amount of law, the analysis can view
how this amount varies with the addition different indicator variables individually or in
different combinations.
19


CHAPTER IV
METHODOLOGY
This research is a quantitative analysis of secondary data. The analysis uses data
from the State Court Processing Statistics; 1990-2006 that includes felony defendants in
large urban counties. The data are expansive and have been consistently collected over
a 16 year time frame with a sample designed to represent large urban areas in all 50
states (Bureau of Justice Statistics). The data set provides a way to capture the drug
arrest, conviction, and penalty information across the Nation (Bureau of Justice
Statistics). Although data are available for 16 years, this research uses three years of
data (2002-2006). The State Court Processing Statistics (SCPS) program tracks felony
cases filed in May of a given year until final disposition or until one year has elapsed
from the date of filing. This data set presents data on felony cases filed in approximately
40 of the nation's 75 most populous counties in even numbered years from 1990-2006.
These 75 counties account for more than a third of the United States population and
approximately half of all reported crimes. The cases from these 40 jurisdictions are
weighted to represent all felony filings during the month of May in the 75 most
populous counties. Data were collected on arrest charges, demographic characteristics,
criminal history, pretrial release and detention, adjudication, and sentencing. Data were
collected in a two-stage sampling process;
20


the first stage was a stratified sample to select 40 of the 75 most populous counties, and
the second stage sample was a systematic sample of defendants based on felony filings
within each selected county (ICPSR 2038). This dataset has many advantages as it gives a
good picture of urban crime, conversely, it fails to capture crime in more rural areas.
Statistical Methods
The independent variables used to describe an offender and their situation that
affects their outcome are: (a) the type of drug crime they are being charged with, (b) the
race of the offender, (c) the offenders' choice or options available for legal
representation, and (d) whether the defendant was released or detained prior to going
to before the judge. The dependent variable is the level/amount of punishment. The
data are split by year and type of drug crime. It is necessary to view minor and major
drug crimes separately as the punishments for these crimes are assumed to be different.
For each level of crime, correlation tables are used to indicate any significant
relationship with each independent variable to the dependent variable (level/amount of
punishment) as well as with each other (Schutt, 2009).
Logistic regression models measures how much each independent variable
(race, type of legal representation, and whether defendant is released or detained) can
predict the final penalty outcome independently. Building on that model, the regression
model measures how much we can predict the penalty outcome by using the entire set
of independent variables (Variable Model, Figure 1).
21


Race of Defendant
Black or White
Drug charge Major or Minor
Private Attorney Public Defender
Detained
Released
Detained
Released
Non-punitive Outcome-
Community Service,
Fine, Restitution,
Treatment, Other
(dismissed or pled out)
Punitive Outcome-
Jail, Probation,
Electronic
monitoring
Non-punitive Outcome-
Community Service,
Fine, Restitution,
Treatment, Other
(dismissed or pled out)
Punitive Outcome-
Jail, Probation,
Electronic
monitoring
Figure 1. Variable Model
To begin the process, the population is divided by minor and major drug crimes.
The difference in the types of crime associated with these two types makes it impossible
to analyze them together. Correlations are used to identify relationships and the
significance of those relationships between race and legal representation, legal
representation and sentencing outcomes, and an additional variable of whether an
individual is released or detained prior to going before a judge (see Appendix A for
operating definitions).
Wealth often determines legal advantage for one individual versus another; to
measure this advantage, the type of legal representation one uses when charged with a
minor or major drug offense is the measure of stratification (Tables 3 and 4). Further,
looking at the relationship between types of representation and whether one is released
22


or detained prior to going before a judge (Table 5 and 6) and then, whether this affects
sentencing outcome shows the relationship and the significance of those relationships
to the dependent variable (penalty or amount of law). To define amount of law, a
collapsed variable of penalty is created that places penalties into two categories;
punitive and non-punitive.
Non-punitive justice will include restorative justice practices that take crime
seriously without increasing repression and exclusion and involved both parties and
focuses on their personal needs. In addition, it provides the offender help to avoid
future offenses. It is based on a theory of justice that considers crime and wrongdoing
to be an offense against an individual or community rather than the state. For the
purposes of this analysis, restorative penalties include community service, fines,
restitution, treatment, and will be included in with dismissals and pleas in the non-
punitive category of punishment.
Punitive justice is the traditional form of justice that is focused on inflicting
punishment and/or a type of revenge. Jail, probation, and electronic monitoring are
included in this category. Logistic regressions will be done to measure the different
relationships to determine which ones are stronger and what is the mediating
circumstance that determines sentencing outcomes. Relationships between variables
and to their significance to the penalty outcome will be analyzed at three time points.
The regressions control for gender and only males are included in the analysis. Ethnicity
is not controlled for so neither race can be considered "non-Hispanic." After controlling
23


for the above areas and selecting the three data points, the dataset has a total N of
10,743 with 3382 in 2002, 3466 in 2004, and 3895 in 2006.
Indicator Variables
For the purposes of this research, the data were coded numerically. The
numbers are not on a scale and there are no meanings, rating, or values assigned to any
of these numbers, they are simply used to analyze the positive and negative
associations. With that understanding, the indicator variables are as follows:
Race Released or Detained?
White-1 Released-1
Black-2 Detained-2
Representation Penalty Received
Public-1 Punitive-1
Private-2 Figure 2. Variable Coding Characteristics of the populations Non-punitive-2
Statistics indicate that the percentages of both White and Black males arrested
for minor drug crimes are similar across all three years of data (Table 1). This fact does
not hold true for major drug crimes. Major drug crimes include drug trafficking, sales,
distribution, possession with intent to distribute or sell, manufacturing, and smuggling
of controlled substances. This category does not include possession of controlled
substances. In 2002, almost three quarters (n=1199, 70.6%) of the major drug crimes
were committed by black males. This difference decreased by 9% in 2004 to 61.6%
24


(n=1046) but rose to over three quarters (n=1282, 75.5%) in 2006 (see Table 2). These
numbers only represent arrests and not convictions for drug crimes.
Table 1. Minor drug crimes by race 2002-2006
White Male Black Male Total
N % N % N %
2002 844 49.9% 847 50.1% 1691 100%
2004 945 47.7% 1035 52.3% 1980 100%
2006 1038 48.2% 1117 51.8% 2155 100%
Data from the State Court Processing Statistics
Table 2. Major drug crimes by race 2002-2006
White Male Black Male Total
N % N % N %
2002 495 29.3% 1196 70.7% 1691 100%
2004 446 30.0% 1040 70.0% 1486 100%
2006 489 28.1% 1251 71.9% 1740 100%
Data from the State Court Processing Statistics
Legal representation for each drug charge type varies but given the very
different nature of each charge, this is to be expected, the crimes included in major
crimes are crimes that involve larger amounts of substance, larger amounts of money,
and more danger. These factors make the stakes higher for the defendant and with
more resources; it may not be unusual for them to hire a private attorney if they can.
The data indicate a difference in legal representation based on race for both minor and
major drug crimes (Tables 3 and 4). Of the White defendants arrested for minor drug
crimes in 2002, 28.9% (n=144) were represented by a private attorney while only 13.2%
(n=78) of Black defendants had that same level of legal representation. This holds true
in subsequent years with 22.1% (n=170) of White defendants and 12.6% (n=98) of Black
25


defendants hiring private attorneys in 2004 and 22.4% (n=163) of White defendants and
11.7% (n=103) of Black defendants hiring private attorneys in 2006. This shows that
over the three time points, White defendants were able to have private legal
representation at a percent double that of Black defendants.
Table 3. Legal representation by race 2002-2006-minor drug crimes
Public Defender Private Attorney
N % N %
White 355 71.1% 144 28.9%
2002 Black 514 83.8% 78 13.2%
White 598 77.9% 170 22.1%
2004 Black 678 87.4% 98 12.6%
White 565 77.6% 163 22.4%
2006 Black 774 88.3% 103 11.7%
As with minor drug crimes, when charged with a major drug crime, White
defendants are far more likely to have a private attorney than their Black counterpart.
Table 4. Legal representation by race 2002-2006- major-drug crimes
Public Defender Private Attorney
N % N %
White 202 59.2% 139 40.8%
2002 Black 639 80.2% 158 19.8%
2004 White 202 66.7% 101 33.3%
Black 594 82.5% 126 17.5%
2006 White 225 64.3% 125 35.7%
Black 773 78.9% 207 21.1%
Data for tables 3 and 4 are from the State Court Processing Statistics
By itself, the difference in legal representation is interesting, however, paired
with a study done by Moyer and Basic (2004) that indicated that detention at the initial
bail hearing disadvantages the individual and their likelihood of taking a guilty plea and
26


receiving a jail sentence it becomes more significant. Tables 5 and 6 represent the
release and detention data for both major and minor drug for all private attorneys and
public defenders represented in this data set. Overall, private attorneys are able to get
their clients released at the initial bail hearing at a higher percentage than public
defenders. In 2002, private attorneys' clients charged for a minor drug crime were
released at the initial bail hearing 81.3% (n=178) of the time and public defenders
clients 59.2% (n=504) of the time (Table 4). This percentage dipped in 2004 to 73.5%
(n=194) released for private attorneys and 51.2% (n=613) for public defenders. In 2006,
this held steady for private attorneys' with 81.4% of their clients being released at the
initial bail hearing, the percentage lowered for public defenders (n=658, 49.7%). These
differences were significant at the .001 level across all three years.
Table 5. Release and detention status by legal representation 2002-2006-minor drug
crimes
Public Defender Private Attorney
N % N %
Released 504 59.2% 178 81.3%
2002
Detained 347 40.8% 41 18.7%
2004 Released 613 51.2% 194 73.5%
Detained 584 48.8% 70 26.5%
Released 658 49.7% 215 81.4%
2006
Detained 665 50.3% 49 18.6%
These differences in release and detention percentages remain very similar for
individuals charged with major drug crimes. In 2002, private attorneys obtained release
for their clients 25.4% more often than public defenders. These differences hold true in
2004 and 2006 with private attorneys obtaining release for their clients more often
27


(27.4% and 26.1% respectively). As with minor drug crimes, these differences are
significant at the .001 level.
Table 6. Release and detention status by legal representation 2002-2006-major drug
crimes
Public Defender Private Attorney
N % N %
Released 462 55.6% 239 81.0%
2002
Detained 369 44.4% 56 19.0%
Released 383 50.1% 176 77.5%
2004
Detained 382 49.9% 51 22.5%
Released 529 53.1% 262 79.2%
2006
Detained 468 46.9% 69 20.8%
Data for tables 5 and 6 come from the State Court Processing Statistics
Overall, Black males are the majority of defendants in all three years of the sample.
They are the most likely to be arrested for a crime, most likely to be represented by a
public defender, and ultimately most likely to be detained prior to going before a judge
for sentencing. This is true in most prison populations and is often what prompts
people to question the fairness of the justice system.
28


CHAPTER V
MINOR DRUG CRIMES
Employing Black's Theory of the Behavior of Law, this research uses the amount
of law as the dependent variable. To define amount of law, a collapsed variable of
penalty was been created that places penalties into 2 categories; punitive and non-
punitive. Using a correlation provides a way to measure how associated or related two
variables are. The research examines the existing data and determines if and in what
way those things are related to each other. Correlations allow prediction's about one
variable based on what we know about another variable.
Correlations
Minor drug crimes involve small amounts of drugs and money and tend to
involve fewer individuals. This type of offense lends itself to a non-punitive approach to
justice that includes more restorative options. Across all three years, the race of the
defendant and legal representation an individual obtains after being charged with a
minor drug crime have a significant negative correlation (2002; r (472) = -.195, pc.OOl.
2004; r (690) =-.101, pc.OOl. 2006: r (847) =-.166, pc.001.). The negative correlation
indicates that Whites, when compared to Blacks are more likely to have private legal
representation. Across all three years, race and whether an individual is detained or
released have no significant relationship. Legal representation and whether one was
29


released or detained prior to going before a judge shows a negative correlation across
all three years (2002; r (472) = -.179, pc.OOl. 2004; 2 (690) = -.261, p<.001. 2006; r (847)
=-.235, p<.001.). This negative relationship shows that as when compared to
representation by a public defender, being represented by a private attorney increases
your chances of being released before going before the judge. For the secondary
variables (not including penalty received), race and legal representation as well as legal
representation and being released or detained showed correlation for all three years of
data. These data are represented in Table 7 below.
Table 7. Variable by variable correlations-minor drug crimes
Legal Released or
Race of defendant representation Detained?
N Pearson Sig- Pearson' Sig- Pearson Sig-
Year 's Corr. (2-tailed) s Corr. (2-tailed) 's Corr. (2-tailed)
Race of Defendant 472 - - -.195** .000 .076** .819
2002 Legal Rep. 472 -.195** .000 - - -.179** .000
Detained or Released? 472 .011 .819 179** .000 - -
Race of Defendant 690 - - -.101** .008 .003 .939
2004 Legal Rep. 690 -.101** .008 - - -.261** .000
Detained or Released? 690 .003 .939 .-261** .000 - -
Race of Defendant 847 - - -.166** .000 -.044** .899
2006 Legal Rep. 847 -.166** .000 - - -.235** 0
Detained or Released? 847 -.004 .899 -.235** .000 - -
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
As shown above, the race variable is significantly correlated to legal
representation across all three years. This trend continued with the penalty received for
30


a minor drug crime. Race and penalty had a negative correlation across all three years
(2002; r (472) = -.303, p<.001. 2004; r (690) = -.208, p<.001. 2006; r (847) =-.124, p<.001)
indicating that as race increases, (up from White-1 towards Black-2) the chances of
receiving a non-punitive justice punishment decreases. Legal representation shows a
positive correlation in 2004 (r (690) = .068, p<.05) indicating that as ones representation
increases, so does the chances of receiving a non-punitive form of justice. Although
some studies have shown that being released or detained prior to going before a judge,
these data did not indicate any correlation between penalties imposed and whether one
is released or detained prior to going before a judge (see Table 8).
Table 8. Variable correlations of independent variables to penalty received for minor
drug crimes 2002-2006
Year Independent Variable Pearson's Correlation Sig. (2-tailed) N
Race of Defendant -.303** .000 472
2002 Legal representation .005 .112 472
Released or detained? -073 .921 472
Race of Defendant -.208** .000 690
2004 Legal representation .068* .003 690
Released or detained? .094 .073 690
Race of Defendant - 124** 0.002 847
2006 Legal representation .042 0.133 847
Released or detained? .018 0.512 847
* Correlation is significant at the 0.05 level (2-tailed).
** Correlation is significant at the 0.01 level (2-tailed).
31


Minor Drug Crimes 2002
In the beginning step of the regression model, with no variables in the model
and no other information about the other variables, the data show that in 2002 if we
just assume that every individual charged would receive a punitive penalty, we would be
correct 55.1% of the time overall (see Table 9). This is significant at the .05 level (Exp.
B=.815, p=.027) (see Table 10). At this point, model 1 currently has a 100% success rate
predicting punitive justice outcomes in 2002. By adding race to the model (model 2),
the ability to predict sentencing outcomes increases by 8.5% in 2002. The logistic
regression indicates that race is significant in 2002 (Exp. B=.323 p=.011) and the
constant or the dependant variable penalty becomes significant at the .01 level (Exp.
B=4.34) and has increased in value indicating that with the addition of race our ability to
predict penalty has become more significant. This is also seen in the Tests of Model
Coefficients (see Table 11) which is also statistically significant for 2002 (X2=35.787,
p=<.01), indicating that by adding race to the model, the increase in the predictability of
the model is significant in 2002. In model 3, legal representation is added. With this
addition, both penalty and race remain significant in 2002 however, by itself, legal
representation is not (Exp. B =.657). Although independently legal representation is not
significant, it does add to the overall model (X2=38.361, p=<.001). In the final model, all
three variables are present; race, legal representation and whether an individual is
released or detained. With all three variables, the predicted percent values remain the
same although the numbers vary slightly. The final model indicates that race and
32


penalty remain significant with each addition to the model and although legal
representation and released or detained were not significant on their own, they added
to the models ability to predict sentencing outcomes. Overall, race is a strong predictor
of sentencing outcomes bin 2002 by itself and as part of the model, legal representation
and released or detained individually are not strong predictors of penalty outcomes
however they do create a stronger predictive model.
Table 9. Predicted and observed values-Minor drug crimes 2002
Observed Punitive Justice Predicted Non-Punitive Justice Percent Correct
Penalty 1-Punitive Justice 260 0 100%
Model 1 O-Non-punitive Overall Percentage 212 0 0.0% 55.1%
Penalty 1-Punitive Justice 161 99 61.9%
Model 2 O-Non-Punitive Overall Percentage 73 139 65.6% 63.6%
Penalty 1-Punitive Justice 161 99 61.9%
Model 3 O-Non-Punitive Overall Percentage 73 139 65.6% 63.6%
Penalty 1-Punitive Justice 165 95 63.5%
Model 4 O-Non-Punitive Overall Percentage 77 135 63.7% 63.6%
33


Table 10. Logistic regression predicting sentencing outcomes-Minor drug crimes 2002
Odds Ratios
Model 1 Model 2 Model 3 Model 4
Penalty only .815* 4.34** 7.842** 393**
Standard Error .093 .298 .479 .198
Race .323** .302** 3.349**
Standard Error .193 .198 .199
Legal Representation .657 .614
Standard Error .264 .270
Released vs. Detained 1.321
Standard Error .209
**Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed).
Table 11. Test of Model Coefficients-Minor drug crimes 2002 X2 df Sig.
Model 1 35.787 1 .000
Model 2 38.361 2 .000
Model 3 40.152 3 .000
Minor Drug Crimes 2004
In the beginning model of 2004, the model shows that if we were to just assume
a sentencing outcome without any other additional information, that guess would be
correct 50% of the time (see Table 12). The constant or the dependant variable
"penalty" in the first model is not significant by itself (see Table 13). When race is added
in Model 2, the penalty variable becomes significant (Exp. B =3.375, p=<.001) and the
race variable is significant (Exp. B= .444, P=<.001). The value of the penalty variable
increases as well as the predictive value of the model (Xz=27.787, p=<.001). Legal
representation is introduced in model 2; this addition had no bearing on the actual
34


predictions and lowered the significance of the penalty variable (Exp. B=2.636, p=<.05)
while the race variable remained significant and saw an increase in its value. Overall,
the addition added to the predictive value of the model (X2=29.495, p=<.001) (See Table
13). In the final model, all three variables are present; race, legal representation and
whether a prisoner is released or detained. With all three variables, the models overall
predicted percentage does not change but the ability to predict each outcome did
change. For example; in model 1, it was predicted that 100% of individuals would
receive a non-punitive form of justice, producing 50% accuracy overall, but this is
reached by predicting 100% correctly for non-punitive outcomes but 0% correct for
punitive outcomes. In model 4, the overall prediction was 50% and the internal accuracy
increased as well (59.4% punitive and 60.6% non-punitive). With all variables in the
model, the overall model fit is significant across models (see Table 13) as in 2002; this
significance indicates that with each variable added to the model, the overall fit of the
predictive model improved. Race has remained significant across models (p=<.001)
although other variables have not. The lack of significance in the penalty variable does not
affect the predictive model, the penalty variable is the dependant variable and does not
affect the other variables but rather is affected by them.
35


Table 12. Predicted and observed values-Minor drug crimes 2004
Predicted
Punitive Non-Punitive Percent
Observed Justice Justice Correct
Penalty 1-Punitive Justice 0 345 0.0%
Model 1 0-Non-Punitive Justice 0 345 100%
Overall Percentage 50.0%
Penalty 1-Punitive Justice 207 138 60.0%
Model 2 0-Non-Punitive Justice 138 207 60.0%
Overall Percentage 60.0%
Penalty 1-Punitive Justice 207 138 60.0%
Model 3 0-Non-Punitive Justice 138 207 60.0%
Overall Percentage 60.0%
Penalty 1-Punitive Justice 205 140 59.4%
Model 4 0-Non-Punitive Justice 136 209 60.6%
Overall Percentage 60.0%
Table 13. Logistic regression predicting sentencing outcomes-Minor drug crimes 2004
Odds Ratios
Model 1 Model 2 Model 3 Model 4
Penalty only 1.000 3.375** 2.363* 1.248
Standard Error .076 .246 .366 .478
Race 444** .453** .454**
Standard Error .155 .156 .157
Legal Representation 1.330 1.507
Standard Error .219 .228
Released vs. Detained 1.399*
Standard Error .163
**Correlation is significant at the 0.01 level (2-tailed).
"Correlation is significant at the 0.05 level (2-tailed).
36


Table 14. Test of Model Coefficients-Minor drug crimes 2004
X2 df Sig.
Model 1 27.787 1 .000
Model 2 29.495 2 .000
Model 3 33.787 3 .000
Minor Drug Crimes 2006
In the last year of analysis, the first model indicates that an initial guess without
any information in the model would yield 57.5% accuracy (see Table 15). At this point
in the model, the constant or dependant variable "penalty" is significant (Exp. B=.739,
p=<.001) (see Table 16). In model 2, race is added to the model, independently, race is
significant (Exp. B=.581, p=<.001) however its addition lowers the significance of the
constant or penalty variable (Exp. B=1.664, p=<.005). While the significance has
decreased, the value has actually increased. The overall model is significant (X2 =15.103,
p-<.001). The addition of legal representation did not increase the overall prediction and
the variable is not significant independently. With the addition of legal representation,
race remains significant and increases in value (Exp. B=.591, p=<.001). The overall
predictive model has increased and continues to be significant (X2=15.602, p=<.001).
The addition of the released or detained variable further lowers the significance the
penalty and legal representation variables although the value of the penalty variable
increased. The race variable remains consistently significant (Exp. B=.585 p=<.001). With
all variables in the model, the overall model fit is significant across all models (Ml;
X2=15.003, p=<.001, M2; X2=15.602, p=<.001, M3; X2=17.310, p=<.001).
37


Table 15. Predicted and observed values Minor Drug Crime 2006
Predicted
Observed Punitive Justice Non-Punitive Justice Percent Correct
Penalty 1-Punitive Justice 487 0 100%
Model 1 O-Non-Punitive Justice 360 0 0.0%
Overall Percentage 57.50%
Penalty 1-Punitive Justice 487 0 100%
Model 2 O-Non-Punitive Justice 360 0 0.0%
Overall Percentage 57.5%
Penalty 1-Punitive Justice 444 43 91.2%
Model 3 O-Non-Punitive Justice 311 49 13.6%
Overall Percentage 58.2%
Penalty 1-Punitive Justice 373 114 76.6%
Model 4 O-Non-Punitive Justice 234 126 35.0%
Overall Percentage 58.9%
Table 16. Logistic regression predicting sentencing outcomes-Minor drug crimes 2006
Odds Ratios
Model 1 Model 2 Model 3 Model 4
Penalty only .739** 1.664* 1.386 1.985
Standard Error .070 .220 .339 .437
Race .581** .591** .585**
Standard Error .140 .142 .143
Legal Representation 1.146 1.077
Standard Error .193 .199
Released vs. Detained .826
Standard Error .147
**Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed).
38


Table 17. Test of Model Coefficients-Minor drug crimes 2006
X2 df Sig.
Model 1 15.103 1 .000
Model 2 15.602 2 .000
Model 3 17.310 3 .001
Across all three years we see that race is consistently a significant predictor of
sentencing outcomes for minor drug crimes. While the other independent variables are
able to add to the model, they are not able to stand on their own or even together
without the existence of the race variable. Donald Black uses race as one of the factors
that determines socioeconomic status and thus affects your vertical status that is
determined by wealth. These factors put together do show a tendency for law to be
applied in greater amounts for minor drug crimes depending on race.
39


CHAPTER VI
MAJOR DRUG CRIMES
Correlations
Major drug crimes involve larger amounts of drugs and money and by default a
greater amount of danger. These offenses often translate into higher penalties for the
defendant charged in these cases. Across all three years, the race of the defendant and
legal representation an individual obtains after being charged with a major drug crime
have a significant negative correlation (2002; r (602) = -.226, p<.001. 2004; 2 (520) = -
.286, p<,001. 2006: r (643) =-.164, p<.001.). Similar minor drug crimes, the negative
correlation indicates that Whites, when compared to Blacks are more likely to have
private legal representation. In 2004 and 2006, race has a positive correlation with
whether a defendant was released or detained (2004; r (520) =.169, p<.001, 2006; r
(643) = .083, p<.005). This positive correlation indicates that Blacks, when compared to
Whites are more likely to be detained before going before a judge or before going to
trial. Legal representation and whether one was released or detained prior to going
before a judge shows a negative correlation across all three years (2002; r (602) = -.268,
pc.001. 2004; 2 (520) = -.306, pc.001. 2006; r (643) =-.267, pc.001.). This negative
relationship shows that as when compared to representation by a public defender,
being represented by a private attorney increases your chances of being released before
40


going before the judge. For the secondary variables (not including penalty received),
race and legal representation as well as being released or detained showed correlation
for all three years of data. Race and being released or detained showed correlation in
2004 and 2006. These data are represented in Table 18 below.
Table 18. Variable by variable correlations-major drug crimes
Race of defendant
Legal
representation
Released or
detained
Year N Pearson's Sig. Pearson's Sig- Pearson's Sig.
Corr. (2-tailed) Corr. (2-tailed) Corr. (2-tailed)
Race of Defendant 604 - - -226** .000 .004 .920
2002 Legal Rep. 604 -.226** .000 - - -268** .000
Detained or Released? 604 .169** .000 -306** .000 - -
Race of Defendant 522 - - -.286** .000 -286** .000
2004 Legal Rep. 522 -286** .000 - - -306** .000
Detained or Released? 522 .169** .000 -.306** .000 - -
Race of Defendant 645 - - -.164** .000 .083* .036
2006 Legal Rep. 645 -.164** .000 - - -267** .000
Detained or Released? 645 .083* .036 -.257** .000 - -
**Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
As shown above, the race variable correlates significantly to both mediating
variables (legal representation and released or detained) in 2004 and 2006. In addition,
race also has a significant relationship with legal representation in 2002. This trend
41


continued with the penalty received for a major drug crime. Race and penalty had a
negative correlation across all three years (2002; r (602) = -.130, p<.001. 2004; r (520) =
-.109, p<.001. 2006; r (643) =-.109, p<.001) indicating that Whites, when compared to
Blacks are more likely to receive a non-punitive justice punishment. Penalty and legal
representation show a positive correlation in both 2002 and 2004 (r (602) = .116,
p<.001; r (520) =.520, p<.001; respectively) indicating that private attorneys, when
compared to public defenders are more likely to obtain a more non-punitive form of
justice. These data did not indicate any correlation between penalties imposed and
being released or detained prior to going before a judge.
Table 19. Variable correlations of independent variables to penalty received for major
drug crimes; 2002-2006
Year Variable Pearson Correlation Sig. (2-tailed) N
Race of Defendant -.130** .001 604
2002 Legal representation .116** .004 604
Released or detained? .009 .833 604
Race of Defendant -.109* .012 522
2004 Legal representation .211** .000 522
Released or detained? -.062 .156 522
Race of Defendant -.109** .006 645
2006 Legal representation .056 .152 645
Released or detained? -.030 .115 645
**Correlation is significant at the 0.01 level (2-tailed).
^Correlation is significant at the 0.05 level (2-tailed).
42


me dUUIUUII Ul edUl VdlldUIC LI IL VUIUL IUI UIL
43


Major Drug Crimes 2002
In the beginning step of the regression model, with no variables in the model and
no other information about the other variables, the data show that in 2002 if we
assumed that every individual charged would receive a punitive penalty, the
assumption would be correct 53.3% of the time. This is not significant (Exp. B=.876,
p=.104) (see Table 21). By adding the race variable race in model 2, the ability to
predict sentencing outcomes increases by 2.7% and the race variable is significant (Exp.
B=.644, p=<.001) independently. With the addition of race, the constant or the
"penalty" variable becomes significant (Exp. B=1.822, p=<.05). The overall model is
significant at the .01 level indicating that race is a produces a significant predictive
model for determining penalty outcomes (see Table 22). In model 3, legal
representation is added to the model. This addition does not change the overall
correct predictions and has lowered the significance of the race variable. Although still
significant, this addition has also lowered the significance of the overall models
predictive value (X2=7.211, p=<.05). The addition of the released or detained variable
raises the prediction minimally (see Table 20). With all of the variables in the model in
2002, the only variable that remains significant is race (Exp. B=.674, p=<.05) although at
a lower significance than at model 2. The overall model is significant but that
significance consistently decreased throughout the model (X2=8.100, p=<.05) and with


drug crimes, the addition of additional variables does not add to our ability to predict
sentencing outcomes.
Table 20. Predicted and observed values Major drug crimes 2002
Predicted
Non-
Observed Punitive Justice Punitive Justice Percent Correct
Penalty 1-Punitive Justice 322 0 100.0%
Model 1 0-Non-Punitive Justice 282 0 0.0%
Overall Percentage 53.3%
Penalty 1-Punitive Justice 230 92 71.4%
Model 2 O-Non-Punitive Justice 174 108 38.3%
Overall Percentage 56.0%
Penalty 1-Punitive Justice 230 92 71.4%
Model 3 O-Non-Punitive Justice 174 108 38.3%
Overall Percentage 56.0%
Penalty 1-Punitive Justice 226 96 70.2%
Model 4 O-Non-Punitive Justice 168 114 40.4%
Overall Percentage 56.3%
Table 21. Logistic regression-Major drug crimes 2002
Odds Ratios
Model 1 Model 2 Model 3 Model 4
Penalty only .876 1.822* 1.385 .810
Standard Error .082 .301 .430 .540
Race .644** .668 .674*
Standard Error .174 .178 .179
Legal Representation 1.185 1.248
Standard Error .191 .199
Released vs. Detained 1.185
Standard Error .180
**Correlation is significant at the 0.01 level (2-tailed).
*Correlation is significant at the 0.05 level (2-tailed).
44


Table 22. Tests of Model Coefficients Major drug crimes 2002
X2 df Sig.
Model 1 6.418 1 .011
Model 2 7.211 2 .027
Model 3 8.100 3 .044
Major Drug Crimes 2004
In 2004 in the beginning model if we were to predict sentencing outcomes with
no information in the model, we would be correct 67.4% of the time (see Table 23). This
is significant (Exp. B=.483, p=<.001). By adding race to the model (model 2), the ability to
predict sentencing outcomes does not change and the constant or dependent variable
"penalty" is no longer significant (see Table 24) however race is significant (Exp. B=.623,
p=<.05). The overall model is significant (X2 =5.482, p=<.019). Legal representation is
introduced in model 3, this does not change the overall percent of correct predictions
however the predictions of individual categories has changed slightly. Legal
representation is significant (Exp. B=2.168, p=<.001) and its addition makes the penalty
variable significant as well (Exp. B=.286, p=<.05) although race is no longer significant at
all. The addition of legal representation added significantly to the overall predictive
model (X2=17.627, p=.001). In the final model and with the addition of the released or
detained variable, all variables are in the model. The percent of correct predictions again
has increased minimally (.6%) and the released or detained variable is not significant.
The only variable remaining significant at this stage of the analysis is legal representation
(Exp. B=2.057, p=<.001). The overall model is significant (X2=18.273, p=<.001). The model
indicates that no one variable can significantly predict sentencing outcomes for major
45


drug crimes in 2004 although legal representation was significant, it was not tested by
itself only in conjunction with race or with race and the released or detained variable.
Table 23. Predicted and observed values-Major drug crimes 2004
Predicted
Observed Punitive Justice Non- Punitive Justice Percent Correct
Penalty 1-Punitive Justice 352 0 100.0%
Model 1 0-Non-Punitive Justice Overall Percentage 170 0 0.0% 67.4%
Penalty 1-Punitive Justice 352 0 100.0%
Model 2 0-Non-Punitive Justice Overall Percentage 170 0 0.0% 67.4%
Penalty 1-Punitive Justice 320 32 90.9%
Model 3 0-Non-Punitive Justice Overall Percentage 137 33 19.4% 67.6%
Penalty 1-Punitive Justice 323 29 91.8%
Model 4 0-Non-Punitive Justice Overall Percentage 137 33 19.4% 68.2%
Table 24. Logistic Regression predicting sentencing outcomes- Major drug crimes 2004 Odds Ratios Model 1 Model 2 Model 3 Model 4
Penalty only .483** 1.076 .286* .375
Standard Error .093 .350 .522 .620
Race .623* .766 .779
Standard Error .201 .212 .214
Legal Representation 2.168** 2.057**
Standard Error .221 .230
Released vs. Detained .847
Standard Error .207
**Correlation is significant at the 0.01 level (2-tailed).
*Correlation is significant at the 0.05 level (2-tailed).
46


Table 25. Test of Model Coefficients-Major drug crimes 2004
X2 df Sig.
Model 1 5.482 1 .019
Model 2 17.627 2 .000
Model 3 18.273 3 .000
Major Drug Crimes 2006
As in previous years; with no additional information in the model beyond the
penalty variable a prediction that 100% of individuals would receive punitive sentencing
outcomes would be correct 57.4% of the time, this is statistically significant (Exp. B =.743,
p=<.001). By adding race to the model, the ability to predict sentencing outcomes does
not change. In model 2 race is introduced into the model however it does not prove to be
significant indicating that by itself, race is not a significant predictor of sentencing
outcomes. In addition, the overall model is not significant (X2=2.619, p=.106). In model 3,
legal representation is added to the model and the overall correct prediction values for
2006 increase slightly (.6%) (See Table 26). With the addition of legal representation the
other variables in the model remain insignificant and legal representation is not
significant when it is interacting with race in the model (Exp. B=1.179) (see Table 27).
The significance of the overall model continues to decrease (X2=3.449, p=.178). In the
final model and the addition of released or detained, all three variables are present. With
all three variables, the overall correct prediction increases only 2.4%. The released or
detained variable is not significant and its addition does not make any other variables
significant. With all variables in the model, the overall model fit (see Table 28) is not
significant in 2006 (X2=4.150, p=.246). This has been a trend throughout the analysis.
47


Major drug crimes show a much different pattern or no pattern at all. There does not
appear to be a consistent predictor of sentencing outcomes that was seen in the minor
drug crime data.
Table 26. Observed and predicted values-Major drug crimes 2006
Predicted
Observed Punitive Justice Non- Punitive Justice Percent Correct
Model 1 Penalty 1-Punitive Justice 370 0 100.0%
0-Non-Punitive Justice 275 0 0.0%
Overall Percentage 57.4%
Model Penalty 1-Punitive Justice 370 0 100%
0-Non-Punitive Justice 275 0 0.0%
Overall Percentage 57.4%
Model 3 Penalty 1-Punitive Justice 337 33 91.1%
0-Non-Punitive Justice 238 37 13.5%
Overall Percentage 58.0%
Model A Penalty 1-Punitive Justice 346 24 93.5%
0-Non-Punitive Justice 246 29 10.5%
Overall Percentage 58.1%
Table 27. Logistic regression predicting sentencing outcomes-Major drug crimes 2006
Odds Ratios
Model 1 Model 2 Model 3 Model 4
Penalty only .743** 1.218 .943 .613
Standard Error .080 .315 .422 .507
Race .750 111 1.289
Standard Error .177 .180 .180
Legal Representation 1.179 1.134
Standard Error .180 .186
Released vs. Detained 1.152
Standard Error .169
**Correlation is significant at the 0.01 level (2-tailed).
Correlation is significant at the 0.05 level (2-tailed).
48


Table 28. Test of Model Coefficients-Major drug crimes 2006
X2 df Sig.
Model 1 2.619 1 .106
Model 2 3.449 2 .178
Model 3 4.150 3 .246
Overall, race remains a significant (p=<.01) predictor for sentencing outcomes in
minor drug crimes. This is both by itself and with the addition of legal representation
and detained or released variables. In major drug crimes, race is not a consistent
predictor. For major drug crimes in 2002 it was significant at the .01 level but that
lowered to .024 with the addition of the legal representation variable and lowered again
to .027 with the addition of the released or detained variable. This trend continues
across years with race losing significance and is not significant at baseline in 2006 for
major drug crimes. This indicates that race has become a weak variable for measuring
penalty outcomes for major drug crimes, though it is a very strong predictor for
measuring sentencing outcomes in minor drug crimes.
49


CHAPTER VII
CONCLUSIONS
In major drug crimes, race and penalty did have a negative correlation across all
three years analyzed. This relationship, however, failed to translate in the predictive
model for major drug crimes and race was not a significant predictor of penalty
outcomes for major drug crimes after 2002. Conversely, race was a significant predictor
for minor drug crimes. This difference can be explained by the varying nature of each
type of crime. Major drug offenses include trafficking, distribution, manufacturing, and
smuggling of controlled substances, all activities that are profitable and far more
dangerous than minor drug crimes. The same qualities that make these crimes more
dangerous allow offenders the opportunity to afford a private attorney. In addition,
major drug crimes have stricter sentencing guidelines than minor drug crimes, leaving
judges with less discretion (Tillyer & Hartley, 2010). Minor drug offenses include lesser
charges such as possession of controlled substances, prescription drug violations, and
possession of drug paraphernalia. These are all crimes that are more harmful to the
victim and indicative of addiction rather than a large drug enterprise.
According to Beckett, Nyrop, and Pfingst (2005), most researchers examining
whether or not arrests represent an accurate measure of unlawful behavior
50


concluded that race plays a comparatively small role in arrests for serious offenses, such
as murder and robbery, but a larger role in the policing and arrests of minor offenses.
This result was seen in the data for major drug crimes. In the minor drug crimes, there
was less longitudinal change than one may expect due to policy and economic changes
nationally. This lack of change may be attributed to the time frame of the data and the
consistency of the political climate and an administration that had other priorities.
Black's theory states that stratification and lower socio-economic status leads to
more upward crime and downward law and the current data for minor drug crimes
supports his position. Other factors may impact arrest and sentencing rates however,
those factors are simply beyond the scope of this research. Changes in the economy
often shift arrest rates, though that shift may or may not close the racial gap in the
prison population. In addition, economic variables create changes in drug policy and
mandatory sentencing structures, which may influence prison populations.
The lack of significance in the major crimes data also supports Black's theory of
stratification when viewed from the standpoint of financial stratification as drug sales
and trafficking are crimes with high financial returns that may even the field between
race as it pertains to financial stratification. Another social factor that could have
bearing on the major crime data for this time frame is crystal methamphetamine.
Crystal methamphetamine became prominent in the early 1990s.
The 2005 National Survey on Drug Use and Health (NSDUH) reported
approximately 1.3 million people were using methamphetamine between the 2004 -
51


2005 and 512,000 self-reported usage within the month prior to being surveyed
(NSDUH, 2006). High usage prompted major crackdowns on the sale and production of
methamphetamine across the country. Unlike many drugs methamphetamine is more
prevalent in White communities, both in use and production. This could be a possible
factor in the lack of result for the race variable in the major crimes data.
In minor drug crimes the results show that the addition of legal representation
and the released or detained prior to going before a judge variable does not mediate
the relationship between race and penalty outcomes. Both variables made a positive
contribution but their presence were not required to affect the penalty outcome.
Without additional years of analysis, meaning cannot be assigned to this finding;
however, it does provide support for Donald Black's theory. Black states that your
position or "vertical layer" in life is largely defined by wealth and race is often a
determining factor when it comes to wealth.
Research into sentencing patterns; particularly for drug crimes that are not only
crimes of poverty but often crimes of addiction, are important to better understand the
failings of the current penal system. According to Cole (2000), $97.5 billion is spent
annually nationwide on the criminal justice system. More than half of these funds are
used for police and prosecution. In contrast, 1.3% of annual criminal justice
expenditures assist with indigent defenses. If drug policies were really designed to keep
people safe, there would be a push for expanded treatment funding dollars,
implementation of treatment on demand legislation, and improved methods for
52


identifying individuals with substance use problems rather than using prison as a
substitute for treatment. Any type of system change would require developing models
of treatment and systems of care that address the different and unique needs of
individuals that are both using substances and involved in the criminal justice system.
Without making these changes, prisons will continue to be crowded with drug users.
That population will not be evenly distributed between races because if treatment is a
requirement for release, and socio-economic status precludes an offender from
receiving treatment, the offender becomes enmeshed in penal institutions. Without a
major systems overhaul in the justice system that accounts for the factors of economic
standing, race, and other individualized issues, the overrepresentation of Black males
being imprisoned for reasons not explained by laws but rather social systems that are
inherently racially biased will continue.
Throughout this research there have been situations brought to the forefront
where discriminatory practices can be acknowledged and rectified or compounded;
primarily when the case is brought before a judge. This is particularly true when it
comes to sentencing. At the time of sentencing, barring any strict sentencing guidelines,
judges have the discretion to assess the case and the person. They can evaluate the
strengths and weaknesses, the support systems already in place, and the individual's
overall ability to participate in a rehabilitative program and then impose an informed
sentence. This discretionary power can do great things. Unfortunately, that power can
be used two ways and judicial discretion can become discriminatory.
53


Limitations
This research shows that race is a strong predictor of sentencing outcomes for
minor but not major drug crimes. As discussed in the conclusions, political and
economic climates can shift trends in drug laws and how they are enforced as well as
how they are adjudicated. For example, this trend is seen in states that began making
the enforcement of marijuana laws a lower priority starting in 2007. To better capture
these shifts and to more accurately understand the predictors of penalty outcomes
future research would be better served with a larger time frame. A time frame that
encompasses different political parties would allow for more historical and contextual
interpretation. Further research into future years will also give perspective into change
and growth over time.
54


APPENDIX A
Operating definitions-
Offender Descriptive Variables
Major Drug offenses-Drug trafficking-lncludes trafficking, sales, distribution, possession
with intent to distribute or sell, manufacturing, and smuggling of controlled substances.
Minor drug offenses-Possession of controlled substances, prescription violations,
possession of drug paraphernalia, and other drug law violations.
Race of Defendant-White male or Black male as identified in their arrest file
Representation/Type of Attorney-l-Public Defender, 2-Private Attorney
Outcome Variables
Released or detained before trail-l-Released, 2-Detained,
Penalties-Most severe sentence received: 1-Jail, 2- Probation, 3 -Community Service, 4-
Fine, 5-Restitution, 6-Electronic Monitoring, 7-Treatment, 8-Other (dismissal or plea).
Penalty recode-for the purpose of logistic regression, these were categorized into
punitive and non-punitive methods of punishment.
Non-punitive-lncludes restorative services (i.e. Community Service, Fines, Restitution,
and Treatment) in addition to dismissals or pleas
Punitive-Jail, Probation, Electronic monitoring
55


REFERENCES
Alexander, M. (2010). The new Jim Crow: Mass incarceration in the age of color
blindness. New York, NY: The New Press.
Beckett, K., Nyrop, K., & Pfingst, L. (2005). Drug use, drug possession arrests, and the
question of race: Lessons from Seattle. Social Problems, 52(3), 419-441.
Black, D. (1976, 2010). The Behavior of Law: Special Edition: Bingley, United Kingdom:
Emerald Group Publishing.
Bobo, L. (2006). Unfair by design. The war on drugs, race, and the legitimacy of the
criminal justice system. Social Research, 73(2), 33.
Brennan, P. K., & Spohn, C. (2008). Race/Ethnicity and sentencing outcomes among drug
offenders in North Carolina. Journal of Contemporary Criminal Justice, 24(4),
371-398. doi: 10.1177/1043986208322712
Chappell, A. T., & Maggard, S. R. (2007). Applying Black's theory of law to crack and
cocaine dispositions. International Journal of Offender Therapy and Comparative
Criminology, 51(3), 264-278.
Cohn, S. F., Barkan, S. E., & Halteman, W. A. (1991). Punitive attitudes toward criminals:
racial consensus or racial conflict? Social Problems, 38(2), 287-296.
56


Cole, D. (2000). No equal justice: Race and class in the American criminal justice system.
New York, NY: The New Press.
Hartley, R. D. (2008). Sentencing reforms and the war on drugs: An analysis of sentence
outcomes for narcotics offenders adjudicated in U.S. district courts on the
southwest border. Journal of Contemporary Criminal Justice, 24(4), 437-461.
Hurwitz, J., & Peffley, M. (2005). Explaining the great racial divide: Perceptions of
fairness in the U.S. criminal justice system. The Journal of Politics, 67(3), 762-783.
Hurwitz, J., & Peffley, M. (2010). Justice in America: The separate realities of Blacks and
Whites. Cambridge Studies in Public Opinion and Political Psychology, 111-139.
Jacques, S. & Wright, R. (2010). Criminology as social control: Discriminatory research &
its role in the reproduction of social inequalities and crime. Crime, Law and Social
Change, 53(4), 383-396.
Johnson, D. (2008). Racial prejudice, perceived injustice, and the Black-White gap in
punitive attitudes. Journal of Criminal Justice, 36(2), 198-206.
Kennedy, R. (1998). Race, crime, and the law. London: Vintage Press.
Lee, R. D., & Rasinski, K. A. (2006). Five grams of coke: Racism, moralism and White
public opinion on sanctions for first time possession. International Journal of
Drug Policy, 17(3), 183-191.
Leech, N., Barrett, K., & Morgan, G.A. (2007). SPSS for Introductory and Intermediate
Statistics: SPSS for Intermediate Statistics: Use and Interpretation, Third Edition.
57


Mahwah, New Jersey: Routledge Academic.
Moore, L. D., & Elkavich, A. (2008). Who's using, and who's doing time: Incarceration,
the war on drugs, and public health. American Journal of Public Health, 98 (1),
S176-180.
Moyer, S. & Basic, M. (2004). Pre-trial detention under the young offenders act: a study
of urban courts. Department of Justice Canada: Youth Justice Research. March
2004. http://www.iustice.RC.ca/eng/pi/rs/rep-rap/2004/rr04 yjl-
rr04 jjl/p23.html#a76
Mustard, D. (2001). Racial, ethnic, and gender disparities in sentencing: evidence from
the U.S. Federal Courts. Journal of Law and Economics 44(1) 285-314.
Radosh, P. F. (2008). War on drugs: gender and race inequities in crime control
strategies. Criminal Justice Studies: A Critical Journal of Crime, Law and Society,
21(2), 167-178.
Phillips, M (2008). Pretrial detention and case outcomes, part 2: Felony cases. CJA New
York City Criminal Justice Agency. New York, NY, March 2008.
Schlesinger, T. (2011). The failure of race neutral policies: How mandatory terms and
sentencing enhancements contribute to mass racialized incarceration. Crime &
Delinquency, 57(1), 56-81.
Schutt, R. K. (2009). Investigating the Social World: The Process and Practice of
Research. Newbury Park, CA: Pine Forge Press.
58


Tillyer, R.( & Hartley, R. D. (2010). Driving racial profiling research forward: Learning
lessons from sentencing research. Journal of Criminal Justice, 38(4), 657-665.
Unnever, J. D. (2008). Two worlds far apart: BlackWhite differences in beliefs about
why African-American men are disproportionately imprisoned. Criminology,
46(2), 511-538.
59


DATA REFERENCES
United States Department of Justice; Office of Justice Programs. Bureau of Justice
Statistics. State Court Processing Statistics, 1990-2006: Felony Defendants in
Large Urban Counties [Computer file]. ICPSR02038-v4. Ann Arbor, Ml: Inter-
university Consortium for Political and Social Research [distributor], 2010-09-03.
doi:10.3886/ICPSR02038
Substance Abuse and Mental Health Services Administration. (2006). Results from the
2005 National Survey on Drug Use and Health: National Findings (Office of
Applied Studies, NSDUH Series H-30, DHHS Publication No. SMA 06-4194).
Rockville, MD.
60