Sociodemographic and legal factors in juvenile justice decision-making


Material Information

Sociodemographic and legal factors in juvenile justice decision-making
Physical Description:
vi, 133 leaves : ; 28 cm.
Potter, Roberto Hugh, 1953-
Publication Date:


Subjects / Keywords:
Juvenile justice, Administration of -- Florida   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1982.
Includes bibliographical references (leaves 116-120).
Statement of Responsibility:
by Roberto Hugh Potter.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000318430
notis - ABU5269
oclc - 09022965
System ID:

Full Text








This section of a dissertation tends to be either abrupt or ver-

bose. I am going to be somewhat abrupt here, and hope those to whom

I owe a great deal know I am grateful.

First, I want to thank the Department of Health and Rehabilita-

tive Services and the intake personnel who cooperated in providing

these data. Without their cooperation I would have had to settle for

doing something else.

Second, a word of appreciation to the entire faculty of the

Sociology Department. There have been some who helped in ways they

don't realize. The clerical staff here has also been of great help

and friendship. The folks at Project Diversion were a great deal of

help and encouragement when we were together, and afterwards. My

many thanks to Sharon Fink for suffering through my handwriting and

typing this dissertation.

Third, my committee members deserve a lot of recognition. They

have put up with a lot of my ignorance, confusion, and despair. In

the process I have learned a great deal. So to Bob, Wilbur, Pam, and

John, my thanks. Special thanks go to Chuck for all the experience,

training, and understanding he has shown me over the last four years.

I hope that I will make him proud of his first product in the Govern-

ment sector.

Fourth, I want to thank my family for their support during my

educational career. There have been times when I really needed

them to keep my eyes focused on my goals and plodding along. My

sisters and their families have been good cushions to fall back on

in times of need. I wish that Margaret E. Potter could have been

around to finally see what a sociologist is, or at least what her

son is.

Finally, my fellow graduate students and friends deserve a

whopping thanks. There have been many fellow students who helped

me sharpen my techniques and knowledge, and for that, I am very

grateful. But most of all, I want to thank them for providing

outlets for frustration, times of laughter, and places to turn

during times of personal disaster. While there have been many

people, I want to thank Susan, Kathy, Jerry, and David for all

their support, friendship, and/or love over the years. Without

them, in particular, I'm not sure I would be writing these ac-




ACKNOWLEDGMENTS............................................... ii

ABSTRACT...................................................... v

ONE INTRODUCTION......................................... 1
Statement of the Problem.......................... 4
Organization of the Present Study.................. 5

ISSUES............................................. 8
Theoretical Approaches ........................... 8
Indicators of Powerlessness....................... 12
Research....................................... 18
Discussion..................................... 42
Summary......................................... 47

THREE DATE AND METHODS..................................... 48
Overview....................................... 48
Data and Sample ................................. 48
The Intake Data Cards............................. 51
Measures, Variables, and Codings.................. 53
Analysis Techniques.............................. 61
Notes............................................ 64

FOUR ANALYSIS OF THE DATA.................................. 65
Introduction................................ 65
Bivariate Analysis................................ 65
Multivariate Analyses............................. 72
Notes.......................................... 89

FIVE DISCUSSION AND CONCLUSIONS ............................ 90
Discussion..................................... 90
Conclusions.................................... 112

BIBLIOGRAPHY................................................. 116

APPENDIX.................................................... 122

BIOGRAPHICAL SKETCH........................................ 133

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




May 1982

Chairman: Charles E. Frazier
Major Department: Sociology

The present study deals with the issue of differential treatment

of offenders in the juvenile justice system. Three levels of juvenile

justice system decision-making are examined, though the primary focus

is upon the intake stage. Concern with this issue grows out of the

theoretical and empirical literature on the effects of sociodemographic

characteristics in justice system decision-making.

Records from one county Health and Rehabilitative Services Intake

unit were utilized, yielding a population of 1,209 cases. Logistic

regression is used to examine the effects of legal and sociodemographic

variables on Intake recommendations and State Attorney's action.

Ordinary least squares regression is used to study judicial decisions.

At the intake stage, legal variables are found to have the greatest

effects. However, age and race have appreciable independent effects

on the likelihood of being recommended for a petition to court. At

the State Attorney stage, legal variables and age are the strongest

predictors. At the judicial stage, legal variables and race show

strong effects on decision-making. Intake recommendation has a sub-

stantial effect on both State Attorney's action and judicial dispo-

sition when it is introduced as an additional independent variable.

These results provide some support for conflict theory and pre-

vious research that indicate that extra-legal factors operate in

justice system decisions. Differential processing on the basis of

age generally is viewed as justifiable, while such processing on the

basis of race is viewed as discriminatory. The condemnation as dis-

criminatory of differential processing on the basis of sociodemo-

graphic factors is viewed as problematic.

The concept of stereotying is used to provide a possible explan-

ation for the results presented here. It is argued that, due to the

volume of cases being processed, intake workers utilize stereotypes

to help guide their decisions. Differential treatment, then, may be

seen to result from the use of stereotypes that are learned in a vari-

ety of situations and that receive reinforcement in many ways.

Stereotyping provides a useful tool to intake workers, and it remains

in use even after patterns of differential processingare pointed out

by outside observers.



The ideal of equal justice under the law is accepted by most

Americans, though perceptions of whether or not equal justice is actu-

ally practiced are mixed. It is not uncommon to hear people say, in

casual conversation about this subject, that who a person is and how

much money and/or power the person has makes a difference in the man-

ner in which the person will be treated in the criminal justice system.

Likewise, it is often pointed out that who a child's parents are,

their income, social standing, and where they live also have effects

on how a child will be handled in the juvenile justice system. The

most common perception is that the higher the social prestige and

socioeconomic status, the more lenient the treatment will be from the

justice system. At the same time, however, people will defend the

justice system as acting in the overall best interests of the society,

and in a generally equitable manner. Thus, an interesting contradic-

tion appears to exist in the popular mind between how the justice

system should work, and how it actually does work.

The general public is not alone in holding such contradictory views.

Sociologically oriented criminologists are also divided over the nature

and meaning of variations in treatment meted out by the criminal and juve-

nile justice systems. On the one hand, the "moral functionalists," as Hagan

(1980) calls them, assert that the law is administered on the basis of

legal considerations, and that nonlegal factors play a relatively minor

and unimportant role. From this perspective little differential treat-

ment or discrimination is expected to exist in the justice systems. On

the other hand, the "moral Marxists" (Hagan, 1980) and the societal re-

action theorists (e.g., Becker, 1973; Schur, 1973) argue that, while

legal factors play the primary role in treatment decisions, nonlegal

factors, particularly sociodemographic characteristics, influence

treatment decisions over and above the impact of legal considerations.

Further, they argue the use of nonlegal characteristics results in dis-

crimination against individuals from relatively powerless social groups.

These divergent perspectives on the operation of the justice systems

will be explored more fully in Chapter Two. For now it is sufficient

to note that for over one-half century this has been a topic of con-

siderable import in the sociological literature.

General and academic interests aside, there is another important

reason for concern over the topic of differential treatment within the

justice systems: Justice system decisions deal with human lives. Ju-

venile justice system decisions may have major effects on young lives.

They have the potential to do great good in a child's life, but they

also have the potential to do much harm. For example, a child who

needs certain types of counseling, protection, or training may receive

help through this system. That attention and help may improve the

child's present condition and future prospects as well. This is an

example of a positive outcome. However, societal reaction theorists

(e.g., Schur, 1973) have argued that being processed through the

juvenile justice system may bring about a negative change in youths'

self-images. This sort of result may increase the probability of fu-

ture delinquent behavior by bringing about circumstances that encour-

age offenders to think of themselves as delinquents. There is evi-

dence to suggest that system contact often does more harm than good

(e.g., President's Task Force on Juvenile Justice, 1967). In addi-

tion to the potential tendency for official processing to do harm as

opposed to good, there is the view among labeling theorists that some

groups are more vulnerable than others to the negative effects. This

is the primary reason for the current interest in differential treat-

ment in both the criminal and the juvenile justice systems.

Exactly what constitutes differential treatment within the jus-

tice system is not clear in recent literature. Treating cases differ-

entially on the basis of legal factors, such as offense seriousness

and prior legal record, is seen to be consistent with a legalistic

view of justice. This view assumes the existence of qualitative dif-

ferences between actors on the basis of legal characteristics. Dif-

ferential treatment of cases on the basis of nonlegal factors, partic-

ularly sociodemographic factors, may be justified in terms of the

general philosophy of the juvenile court and the specific principles

of individualized justice. Consistent differential treatment of a

particular social category, especially when the category involves

groups considered to be generally disadvantaged, could be interpreted

as discrimination. It is at least logical to say that differential

treatment indicates a flaw in the operation of the American justice

system. As will be pointed out in Chapter Two, researchers differ

widely in their interpretations of the existing empirical literature,

so that even when evidence of differential treatment is clear, there

is still the matter of deciding whether or not it constitutes discrim-

ination. The issues surrounding this topic are complex and unresolved,

and therefore remain interesting and problematic to sociological re-


Statement of the Problem

The present study is concerned with the issue of differential

treatment of children who enter the juvenile justice system. It will

focus on processing decisions made at the intake stage of the juvenile

justice system. Intake represents an intermediate step between arrest

and prosecution. The mandate for intake workers involves the review

of juveniles' cases shortly after their arrest and then making a rec-

ommendation to the State Attorney's office on whether or not the

child's case should be petitioned to court. Intake decisions are

simply recommendations to the State Attorney; they have no binding

power. They are important, however, because they represent intake

workers' perceptions of qualitative differences between cases recom-

mended for "petition" and for "no petition." Further, State Attorneys

use the intake report to make their decision on whether or not to

prosecute the case. The perceptions and recommendations of intake

workers may impact greatly on later case processing decisions. In

addition, the same agency charged with the intake process also handles

nearly all treatments that may be accorded to juveniles by the court.

Because of this connection, impressions and recommendations made at

the front-end of the process may continue to have an influence through-

out all stages of treatment. The intake process and its role in the

juvenile justice system will be discussed in more detail in Chapter


Previous studies of juvenile justice system decision-making have

drawn on various theoretical perspectives. They have utilized many

different analytic techniques, and not surprisingly, they have presented

inconsistent results. These previous studies will be examined and dis-

cussed in Chapter Three. Like previous studies (Horowitz and Wasserman,

1980), this study will be concerned with two major foci: 1) the deter-

mination of what legal and nonlegal factors are related to juvenile

justice decision-making; and 2) whether such factors are used in a man-

ner that affects juveniles from relatively powerless social groups in

an adverse manner. In specific, the present study will attempt to

determine whether sociodemographic factors such as age, sex, and race,

play important roles in decision-making after the effects of legal

case characteristics have been taken into account. To address this

question, data have been drawn from the intake records of one Florida

county over an 18-month time period. These data contain relevant in-

formation on both legal and sociodemographic characteristics of the

juveniles and their cases. The analytic techniques utilized to study

the issue range from bivariate crosstabular analysis to logistic re-

gression techniques.

Organization of the Present Study

In order to study the issue of differential treatment at the in-

take stage of the juvenile justice system, several steps had to be

taken. These included reviewing previous theoretical and empirical

work on this subject, data gathering and analysis, and interpreting

the results. Chapter Two presents the major theoretical and empiri-

cal literature on differential treatment and indicators of powerless-

ness. Reviews of previous research on differential treatment at

various stages in both the criminal and the juvenile justice systems

are presented. Particular attention is given to previous studies of

juvenile court decision-making.

Chapter Three deals with the data and methods used in this study.

Data collection is discussed first and a brief outline of the intake

process follows. The steps that were taken to turn information ob-

tained from the data source into data for the present analysis also

are described. Finally, the statistical techniques employed in the

data analysis are described and discussed.

Chapter Four presents the results of the analyses of these data.

It proceeds through three steps, ending with a multivariate analysis.

The results of the analysis of intake recommendation is first. The

results of prior research and the criticisms of that research are con-

sidered in relation to the present findings. The effects of the inde-

pendent variables (those used to estimate predictions of intake recom-

mendation) and of intake recommendations on both State Attorney's

action and the final judicial disposition are also examined.

Then, Chapter Five discusses the results in light of the theoreti-

cal approaches outlined in Chapter Two. There is also a critique of

the methodologies used in previous research and this study. Possible

explanations for the observed results are offered. Finally, the


findings from the present study are discussed both in terms of the

practical and the theoretical considerations they raise.




Theoretical Approaches

For over one-half century some sociological criminologists have

been charging that differential processing on the basis of "legally

irrelevant" factors exists in the criminal justice system. Charges

have come from both conflict and societal reaction theorists. These

theorists share a concern with the concepts of power and conflict as

they relate to deviance definition, processing, and labeling. While

the exact explanation of the differential treatment, as well as its

genesis and effects, varies somewhat from writer to writer, these the-

orists agree that processing influenced by nonlegal factors has no

place in the criminal justice system.

Earlier American conflict criminologists such as Sellin (1938)

and Miller (1958) viewed the conflict between cultural and group or

class norms as the sources of definitions of criminal behavior. In

these formulations, the norms and values of the more powerful group

are used to judge the behaviors of the less powerful groupss. Be-

haviors that differ from those of the dominant group are more likely

to be defined as criminal. More recent conflict criminologists have

stressed the importance of political and economic power in the formu-

lation and application of criminal law. They have argued that

criminal laws are formulated and applied to protect the hegemony of

the dominant group or class (Turk, 1969; Chambliss, 1969; Chambliss

and Seidman, 1971; Quinney, 1975, 1977). These later theorists do

not view definitions of criminal behavior as being based on value

conflict as such, but rather as being based on the efforts of more

powerful groups to secure their economic and political interests.

Regardless of the particular conflict theorist, the concept of

power occupies the central role in their analysis of the process of

criminalization, that is, the process of being defined as a criminal.

Later theroists argue that the behaviors of the less powerful groups

or classes are most subject to being defined as criminal, and that

their actual behavior is more likely than that of the powerful class-

es to be scrutinized by control agents. This heightened scrutiny

increases the likelihood that they will be detected performing crim-

inal activities, will be arrested, and will receive harsher treatment

at each stage in the criminal justice system (Chambliss, 1969; Turk,

1969; Chambliss and Seidman, 1971; Quinney, 1975, 1977).

Societal reaction theorists are in agreement with the conflict

theorists about the role of power in the creation of criminality. As

Becker (1973: 17-18) notes: "Differences in the ability to make rules

and apply them to other people are essentially power differentials.

Those groups whose social position gives them weapons and power are

best able to enforce their rules." Power, then, occupies a place of

central importance for several theorists in their analysis of reac-

tions to criminal behavior (Lemert, 1951; Scheff, 1966; Rubington and

Weinberg, 1968; Schur, 1971, 1973; Becker, 1973; Hawkins and Tiedeman,

1975). Societal reaction theorists are concerned with the relationship

between powerlessness and an individual's susceptibility to labeling.

Conflict criminologists tend to direct their attention to the im-

pact of one's position in the social structure as it affects one's

power to influence and cope with the legal machinery. When predicting

susceptibility to criminalization, they do not concentrate on the

identity of the individual, but rather on the group or social category

into which he or she fits. Conflict theorists, particularly the more

recent ones, argue that the behaviors of the less powerful groups or

classes are most subject to being defined as criminal by the law, and

that the actual behavior of the members of these groups is more likely

to be scrutinized by control agents than is the behavior of the more

powerful classes. These theorists, for the most part, confine their

analyses to the effects of structural variables, though they do specify

the correlates of social position that indicate relative powerlessness.

Societal reaction theorists, as pointed out above, tend to focus upon

the interactional level and the effect that possessing one or more of

these structural indicators has on the individual coming into contact

with the criminal justice system. In so far as the conflict theorists

and societal reaction theorists agree upon the symbols and attributes

of powerlessness, these two theories are fairly consistent in their

analysis of how certain groups tend to be handicapped in the criminal

justice system and on what types of differential processing should be

expected when processing decisions are examined.

Socioeconomic status (SES) is the indicator of powerlessness that

has been most often cited by these writers (Lemert, 1951; Chambliss,

1969; Turk, 1969; Quinney, 1970; Chambliss and Seidman, 1971; Schur,

1971, 1973). Chambliss (1969) and Chambliss and Seidman (1971) con-

centrate on SES as the major indicator of powerlessness. Other writ-

ers have broadened the range of factors that are viewed as indicators

of powerlessness. Both Lemert (1951) and Turk (1969) have listed the

social distance between the accused and the social control agent as

an indicator of relative power. The greater the gap, in the favor

of the control agent, the more likely there will be a sanction against

the accused. Turk (1969: 17-18) lists other factors as well as those

mentioned above: minority group membership, youth (aged 16 to 24 years),

being male, being transient, and living in an urban area. Persons from

these categories, particularly those who fit into several categories,

are most likely to be criminalized.

Quinney (1970: 213-227) is in basic agreement with Turk (1969),

asserting that younger people, males, lower-SES persons, and ethnic

minority members are more likely than others to be arrested and pro-

cessed for behaviors that have been defined as criminal. Schur (1973:

125-126), writing primarily about the juvenile justice system, notes

that lower-class, black, and male children are more likely than their

counterparts to be subject to definitions of delinquent behavior.

Finally, Becker (1973: 17, 204) argues that women, blacks, ethnic

minorities, and the lower classes have their rules made for them by

other, more powerful groups and are more subject to being labeled as

criminals. The reasons that these sociodemographic factors are con-

sidered to be important indicators of powerlessness will be explored

in greater detail below.

Indicators of Powerlessness

SES, as mentioned above, has been considered by some to be the

overarching indicator of powerlessness (Chambliss, 1969; Chambliss and

Seidman, 1971; Quinney, 1977). For other conflict and societal reac-

tion theorists, SES is one of several dimensions of powerlessness (Turk,

1969; Quinney, 1970, 1975; Schur, 1971, 1973; Becker, 1973). Among

the later conflict theorists, the effects of SES are explicitly out-

lined, whereas among the earlier theorists (e.g., Sellin, 1938; Miller,

1958) its effects are left somewhat unspecified. A person's SES, com-

posed of such factors as income, education, and occupation, provides

cues about his or her position in the social structure. This structural

position affects one's ability to command economic and political re-

sources that will determine not only one's likelihood of coming under

the scrutiny of social control agents, but also one's experience in the

criminal justice system (Chambliss and Seidman, 1971). Those in lower-

SES categories are more closely scrutinized by social control agents,

less able to avoid arrest for criminal behavior, and less able to cir-

cumvent subsequent legal processing than are those from higher-SES

categories. Location within the social structure, as it affects one's

resources and power, is of crucial importance in the process of label-

ing within the criminal justice system.

Distinctions between race and ethnicity often become blurred in

theory and research on criminal justice processing. This problem may

be avoided by using the term minority group status, which allows the

integration of earlier and more recent conflict theories, as well as

societal reaction theories. The high correlation among minority

status, SES, and social prestige has made minority status an impor-

tant variable for all of these theorists. Whites in American society

generally occupy the most powerful and dominant social positions.

Blacks, Hispanics, Native Americans, and Orientals occupy the less

powerful positions. Blacks have generally been considered the least

powerful minority category in American society, though they are possi-

bly the largest minority group. They also tend to be the most often

criminalized group.

With the exception of Becker (1973), most conflict and societal

reaction theorists view males as the most susceptible to criminaliza-

tion of the two sexes. It should be pointed out, however, that these

other theorists are addressing primarily criminal behavior, while

Becker (1973) is addressing the general area of deviance. Sex thus

seems to play a paradoxical role in these analyses of the criminal

justice process. Lemert (1951), Turk (1969), and Quinney (1970) note

that males are the most often criminalized of the sexes, particularly

younger males (between 16 and 24 years of age), and conclude that be-

ing male is worse than being female if one is being processed through

the criminal justice system. Thus, while women are generally consid-

ered the powerless sex, they are also the least likely to be criminal-

ized. Quinney (1970: 216) argues that this seeming discrepancy is re-

solved by considering the role expectations for men and women. Men

tend to be involved in more aggressive roles, and thereby are more ac-

tive. This activity may lead to more behaviors that have a high prob-

ability of being defined as criminal. Women, on the other hand, are

expected to be more passive, and thereby avoid activities that

traditionally are defined as criminal. Males are therefore more sus-

ceptible to criminalization than are women. Role expectations, as

they relate to social position and behaviors, play an important part

in whether men or women will be likely to engage in activities that

are potentially criminal.

Lemert (1951), Turk (1969), Quinney (1970), Schur (1971, 1973),

and Becker (1973) all argue that the older adolescents and young

adults are in the age category most likely to be criminalized. Quin-

ney (1970: 214) notes that persons in this age group tend to be ar-

rested disproportionately for violent and property crimes. Quinney

(1970: 215) further asserts that much of the traditional role-estab-

lishing behavior of youth has been subjected to criminal definition

because of their relatively powerless position in the social structure.

That is, behaviors that young people feel are necessary in order for

them to establish an adult identity and to achieve the symbols that

convey adulthood are defined as criminal by the older, more powerful

segments of society. One's position in the age structure of society

thus has an impact on how much power is available to avoid definition

as a criminal for certain behaviors.

It should be pointed out that none of these factors are seen to

operate independently of the others in determining the probability of

criminalization or the ability of individuals to avoid a criminal

label. Nor should only these factors be examined when studying the

social correlates of powerlessness. Other indicators have been listed

above and should be kept in mind when studying processing in the crim-

inal justice system. Combining these factors may help to place an

individual at a particular position in the social structure, thereby

allowing a more accurate prediction of the probability of engaging in

behaviors that are definable as criminal, of being observed and arrest-

ed, and of being processed through the criminal justice system. This

is important since, according to conflict and societal reaction theor-

ists, it is the individual's position within the social structure, and

the attributes associated with that position, that determine one's

probability of being criminalized.

Charges of differential processing have not been limited to cri-

tiques of the adult criminal justice system. Indeed, such accusations

are possibly even more numerous when the juvenile justice system is

under scrutiny. Juvenile justice's basis in the concepts of parents

patriae (the state as the ultimate parent) and "individualized justice"

(the philosophy of "in the best interests of the child") have opened

the way for criticism almost since its inception. The coupling of

these two concepts, in part, has created a system with almost unbri-

dled discretion. This broad discretion has come under attack, largely

because of the emphasis placed on the biography and character of the

child rather than on purely legal considerations. The inclusion of

nonlegal factors as major considerations in treatment decisions has

led to charges that such factors are sometimes used in a manner that

biases the outcome against those from the less powerful segments of


Platt (1977) and Krisberg and Austin (1978) have leveled charges

of class bias against the juvenile justice system. Platt (1977) has

provided the most in-depth analysis of the class-based interests that

led to the development of the juvenile justice system. He contends

that the juvenile court movement represented an attempt, both con-

sciously and unconsciously, by the middle- and upper-classes of the

time to control the behavior of the lower-class youth. The invention

of compulsory public education, along with this embodiment of stan-

dards of conduct for children, was to serve as a means of training

lower-class youth for the industrial workplace. Control over the

behavior of the youth from the lower-class also helped to cut down

on outward displays of dissatisfaction that might threaten the social

and economic positions of the middle- and upper-classes. According

to Platt (1977), the juvenile court's lack of concern with procedural

rights and its emphasis on individualized justice and discretionary

decision-making led to sanctioning that was based on the attributes

of the child. Since middle- and upper-class standards of behavior

were imposed on lower-class youth, Platt (1977) contends that lower-

class youth were penalized from the start of their contact with the

juvenile justice system. This bias is still built into the juvenile

justice system today, since the behaviors of lower-class youth are

more likely to bring them into contact with the juvenile justice sys-

tem (Platt, 1977; Krisberg and Austin, 1978). The important point for

these authors, however, is that the standards of behavior embodied in

the juvenile law are basically aimed at controlling behaviors most com-

mon to the less powerful segments of society.

Krisberg and Austin (1978) extend their analysis beyond the class

argument to charge that juvenile laws and procedures were and are aimed

at even more specific targets within the lower-classes. For example,

they illustrate that juvenile laws have been used to harass black

youths active in the civil rights movement in an attempt to maintain

racial boundaries (Krisberg and Austin, 1978). Chesney-Lind (1977),

Schlossman andWallach (1978), as well as Krisberg and Austin (1978),

argue that the juvenile justice system is used to reinforce conven-

tional sex roles and behavioral ideals. These authors contend that

juvenile laws serve to punish girls for behavior that is not consid-

ered proper for females. They charge that sexual discrimination is

built into the juvenile justice system. Females are prosecuted dispro-

portionately, and more severely, for "status" and "decorum" offenses

that are not perceived to be consistent with "proper" female behaviors.

While Platt (1977), Chesney-Lind (1977), Krisberg and Austin

(1978), and Schlossman and Wallach (1978) may all be classified as

conflict theorists, not all such criticisms of the juvenile justice

system have come from conflict theorists. Those associated with the

societal reaction approach have also been critical of the manner in

which nonlegal factors are used in the juvenile justice system, par-

ticularly as they impact on the less powerful social categories. For

example, both Emerson (1969) and Cicourel (1968) note the use of

"moral character" and other factors associated with the biography of

the child as determining factors in juvenile justice decision-making.

Variables such as family status, school behavior, neighborhood, and

attitude are used to construct an image of the youth upon which deci-

sions are based. This image, along with legal factors such as offense

seriousness and prior legal record, is then used to help determine

probabilities of future delinquent involvement. On the basis of

these considerations, processing and treatment decisions are made.

Cicourel (1968) argues that the "background assumptions" carried

by juvenile justice workers contain certain "recipes" for action, be-

cause they carry implicit judgements about the character of the ju-

venile and the nature of the offense. These considerations influence

workers' assessments of the delinquent proclivities of the juveniles.

In a similar vein, Schur (1973) discusses the use of stereotyping for

juvenile justice decision-making. He contends that juvenile justice

decision makers base many of their decisions on stereotypes about the

social and personal, as well as legal, attributes of the juvenile.

Through these stereotypes, which may or may not be true in the case of

the particular child, assumptions are made that affect processing and

treatment decisions. Generally, it is argued, these stereotyped

images work to the disadvantage of the juveniles from less powerful

social groups. In short, analysts from the societal reaction perspec-

tive also accuse the juvenile justice system of decision-making on the

basis of nonlegal factors that tend to work to the disadvantage of the

less powerful social categories.


The theoretical literature outlined above predicts that differ-

ential processing on the basis of nonlegal factors such as SES, race,

sex, and age will be found in the juvenile justice system. While no

one seems willing to argue that such factors are not important in

juvenile justice decision-making, there are those who question the

charges that they are used in a discriminatory manner (e.g., Cohen and

Kluegel, 1978, 1979a, 1979b). Not only have academic researchers ques-

tioned this assertion, but many practitioners also take exception to

these charges. Consequently, there has been a plethora of research

studies aimed at determining exactly what factors are used in juvenile

justice decision-making and whether the use of nonlegal factors is re-

lated to harsher treatment for certain groups.

For many of the same reasons, and with a longer history, these

questions have been raised and explored in the criminal justice system.

This research has served to set the stage for many of the studies of

juvenile justice decision-making. It has influenced the phrasing of

research questions, the methodological designs, and interpretations of

the results. Because of the similarities between studies of the crim-

inal and the juvenile justice systems, and because of the more detailed

critiques of research on the criminal justice system, the review of the

empirical literature begins with an analysis of studies of decision-

making in the criminal justice system.

Differential Processing in the Criminal Justice System

The empirical research on processing decisions in the criminal

justice system, despite the theoretical predictions outlined above,

has failed to show consistent evidence of processing decisions based

on nonlegal factors. Review essays of the research literature deal-

ing with differential processing in the criminal justice system have

repeatedly pointed out the inconsistencies and ambiguities of the re-

sults (Hindelang, 1969; Hagan, 1974; Wellford, 1975; Williams, 1980).

Williams (1980) reports that of 89 studies concentrating on differ-

ential processing in either the criminal or juvenile justice system,

half found evidence of race or SES bias, while the other half did not.

Wellford (1975: 343), reviewing the literature dealing directly with

assumptions taken from a societal reaction perspective, concludes

that there is no empirical evidence of differential labeling on the

basis of nonlegal factors in the criminal justice system. Similarly,

in his review of the sentencing literature, Hagan (1974: 379) con-


review of the data from twenty studies of judicial
sentencing indicates that, while there may be evi-
dence of differential sentencing, knowledge of
extra-legal offender characteristics contributes
relatively little to our ability to predict judi-
cial dispositions. Only in rare instances did
knowledge of extra-legal attributes of the of-
fender increase our accuracy in predicting judi-
cial disposition by more than five percent.

Researchers frequently cite methodological problems when attempt-

ing to explain the reasons for these ambiguous findings. For example,

Hindelang (1969: 312-313) lists four basic problems with the pre-1965

research that found race differentials in sentencing: (1) their data

were drawn from southern states; (2) they failed to control for rele-

vant nonracial factors; (3) their data tended to be more than 10 years

old (1955 or older); and (4) they dealt with capital offenses only.

Williams (1980: 22) adds further methodological considerations in his

characterization of the literature on racial differentials in process-


Upon examination of the study characteristics, some
consistencies emerge. It seems that those studies
which find racial differentials in processing are
more likely to be: at later stages in the system
(sentencing and postsentencing); of adult popula-
tions; of capital offenses; from southern juris-
dictions; older, with a least half the data coming
from populations prior to 1961; and more simplistic
in their data analysis. As a cautionary note, how-
ever, these characteristics should not be taken as
exclusive. Studies with the same characteristics
can, and do, produce opposite findings.

On the other hand, of those studies which do not
produce findings of racial processing differen-
tials, characteristics tend to be almost the op-
posite of those above: earlier stages in the sys-
tem; juvenile populations; noncapital offenses,
western or non-US jurisdictions; samples with
median years of 1961 or after; and more rigorous
analyses. Socioeconomic studies lacking support
for the existence of processing differentials
are of a similar nature but the area of the sys-
tem and juvenile/adult populations provide little

More recent research on differential processing generally has

employed more sophisticated methodologies. Several of these studies

have found that the race of the offender does have a substantive im-

pact on the type of disposition received (Swigert and Farrell, 1977;

Lizotte, 1978; Unnever et al., 1980). Still other have found that

SES has an effect on the disposition accorded a defendant (Farrell,

1972; Burke and Turk, 1975; Swigert and Farrell, 1977; Jacobs, 1978;

Lizotte, 1978; Unnever et al., 1980).

Other nonlegal factors also have been found to be significant

predictors of dispositions received at various stages of the adult

justice system. For example, the sociodemographic variables age

(Burke and Turk, 1975; Bernstein et al., 1977a; Unnever et al., 1980),

sex (Bernstein et al., 1977a), occupation status (Burke and Turk,

1975; Lizotte, 1978), and economic power (Jacobs, 1978) have been

found to be important predictors of sanction severity. Further, per-

sonal characteristics such as demeanor (Hagan, 1975a; Frazier et al.,

1980) and physical appearance (Frazier et al., 1980), have been found

to significantly impact on decisions made in the criminal justice sys-


Researchers, however, still find it difficult to know whether

their findings are indeed evidence of differential treatment in crim-

inal justice processing. Generally, they must work with data sets

that limit the types and number of control variables that can be in-

troduced into their analyses. Controls for variables such as offense

seriousness and prior legal record have been found to make zero-order

relationships between SES, race, and processing outcomes disappear

(e.g., Willick et al., 1975; Cohen and Kluegel, 1979b). Thus, while

some studies utilizing sophisticated analytical techniques have found

nonlegal factors to be related to dispositions, other similar studies

have found legal factors to be the only important predictors of dis-

positions (e.g., Chiricos and Waldo, 1972; Hagan, 1975b; Perry, 1977).

Even when relationships between nonlegal variables and processing de-

cisions appear stable, that is, when they cannot be controlled away,

it is difficult to know whether they are large enough to be of sub-

stantive importance. Researchers tend to rely on statistical signif-

icance to judge the effects of variables, and this has caused at

least one reviewer (Hagan, 1974) to question the substantive impact

that nonlegal variables have on dispositions. It should be empha-

sized that legal case characteristics are found to be the most

important factors in studies of criminal justice processing regardless

of the stage of the system under scrutiny or the level of methodologi-

cal sophistication employed by researchers.

In short, there does not seem to be a great deal of consistency

in the results of studies dealing with differential processing in the

criminal justice system. Findings have varied depending on such fac-

tors as the region of the country and the historical period from which

the data were collected. Any summary about the role that nonlegal

factors play in criminal justice decision-making processes must be

tempered by the fact that current research is, at best, inconclusive.

More rigorous empirical studies are needed to explore this issue.

Emphasis needs to be placed on the comparability of measures, new

data, and improved methodologies.

Differential Processing in the Juvenile Justice System

Empirical research conducted on the question of differential pro-

cessing in the juvenile justice system does not escape the criticisms

leveled against such research in the criminal justice system. In the

following pages, examples of empirical research on several juvenile

justice decision-points will be examined.

Review essays on differential treatment in the juvenile justice

system reach varying conclusions about the outcome of the empirical

research (Barton, 1976; Liska and Tausig, 1979). Barton (1976: 478)

concludes that:

it appears that factors other than the present offense
take on increasing importance in determining the fate
of a youth as he becomes more involved in the juvenile
justice system, probably because those responsible for

determining his fate have available increasing amounts
of information about him as he penetrates further into
the system.

He goes on to note that girls tend to fare more poorly than boys in

this system, but that the effects of being black versus white are not

as clear-cut. He suggests that the correlation between being black

and other factors, such as family status, school achievement, and

social class, may work to the detriment of black youth in the system.

Liska and Tausig (1979: 199), on the other hand, argue that there

is little empirical support for the contention that lower-SES youths

are treated more harshly in the juvenile justice system. By contrast,

however, they point out that being a minority group member has been

shown to increase a juvenile's chances of severe sanctioning.

Obviously, these two review studies disagree about the relative

impact of nonlegal and legal factors in juvenile justice decision-mak-

ing. One reason for this disagreement may lie in the fact that much

of the research cited by Liska and Tausig (1979) was conducted after

Barton's (1976) article. An in-depth review of the literature dealing

with juvenile justice processing decisions may help clarify some of the

major issues addressed in these review essays.

Arrest decisions

The study of differential processing generally begins at the level

of police-citizen contact because it is at this point that a person

first comes into contact with law enforcement officials. Much of the

research on police-citizen contacts has been conducted on police con-

tacts with juveniles. Decisions to confront and/or arrest a juvenile

(or an adult) are not simple. As the review essay by Thomas and Fitch

(1977) points out, many factors come into play when an officer is in

the field. Studying this issue necessarily involves understanding

these factors. The methodologies employed in research on police level

decision-making range from participant observation, i.e., "ride along"

methods (Piliavin and Briar, 1964; Black and Reiss, 1970), to examina-

tions of official records (Weiner and Willie, 1971) and survey data

(Ferninand and Luchterhand, 1970; Williams and Gold, 1972). Each

method has drawbacks, but each also provides an interesting view of

the process of contact and arrest. Thomas and Fitch (1977) note, in

their review of police-citizen interactions, that police contact and

arrest decisions are generally reactive, influenced not only by legal

factors, but also by personal characteristics of the offender and the

victim, as well as the situation in which the offense occurs.

Piliavin and Briar's (1964) observation of police encounters with

juveniles emphasized the limited role that legal factors seem to play

in decisions to further process an apprehended juvenile. They noted

only one legal factor, prior record, to be related to the decision to

arrest a youth. Black and Reiss (1970) observed that the attitude of

the victim (i.e., understanding and lenient versus retaliatory) played

a major role in the officers' decisions to arrest a juvenile. Since

most police contacts with juveniles are reactions to citizen complaints,

rather than selections of youth by police officers, Black and Reiss

(1970) argue that personal characteristics of offenders play a rela-

tively small part in police-decisions. The probability of arrest also

increased with the severity of the offense, although only about 5

per cent of the alleged crimes during the study period were felony of-


Overall, these two studies suggest that factors not directly re-

lated to the offense play a substantive role in decisions made at this

stage of the juvenile justice process. Some authors have examined the

direct effects of traditional sociodemographic variables (e.g., race,

sex, SES, and age). Two studies (Piliavin and Briar, 1964; Weiner and

Willie, 1971) have noted that blacks tend to be contacted by the police

more often than whites. Weiner and Willie (1971) and Williams and Gold

(1972) both reported an association between police contact and SES;

lower-class juveniles were contacted most often by the police. Older

juveniles have also been observed to have a greater probability of

contact/arrest than younger children (Piliavin and Briar, 1964; Wil-

liams and Gold, 1972). Sex of the juvenile has not been shown to have

any significant association with contact/arrest decisions, largely

because most studies do not address the issue of gender and police


Demeanor is another nonlegal factor that has been found to be

related to arrest decisions for juveniles. A youth's demeanor has

been an important variable in observational studies (Piliavin and

Briar, 1964; Black and Reiss, 1970), as well as in one survey study

(Ferdinand and Luchterhand, 1970). Because demeanor may be associated

with factors such as race, sex, SES, and neighborhood (Piliavin and

Briar, 1964), it may be one way that race or SES differentials are

introduced into arrest patterns. Youths who act "tough" exhibit poor

demeanor and may be more likely to come from lower-SES and minority

groups (Piliavin and Briar, 1964; Black and Reiss, 1970).

In sum, sociodemographic and other nonlegal factors have been

found to be important in police-juvenile contacts and arrest decisions.

The use of nonlegal factors at this decision point generally works to

the disadvantage of lower-SES and minority group youth. Differentials

in the manner in which police apply the law may operate to reduce the

variance in the population of youth that reaches each successive stage

in the juvenile justice system. Independent of this issue, however,

the focus of this review now shifts to decisions made about delinquency

cases at later stages of the juvenile justice system.

Intake decisions

Normally the next step after arrest in the juvenile justice pro-

cess is the intake stage. Literature on this stage of the system will

be examined in somewhat more depth than other stages because it is the

major focus of the present study. Perhaps the earliest major study of

intake decisions was conducted by Goldman (1969). The police depart-

ment in the area from which he gathered his data was responsible for

intake. Goldman noted that only 35.4 percent of the juveniles con-

tacted by the police were referred to court. Of these, the more seri-

ous the offense for which the youth was apprehended, the more likely

that a petition to court would be requested. He also found that black

youths were referred to court more often than white youth, for both

major and minor infractions. While not a statistically significant

relationship, Goldman reported that girls were more likely than boys

to be referred to court. Finally, as the age of the juvenile increased,

so did the likelihood that a court petition would be requested by the


McEachern and Bauzer (1967) studied police records for two Cali-

fornia cities to determine what factors influenced police to request

court petitions. In one city, age of the juvenile, police precinct,

offense seriousness, prior record, and probation status were found to

be related to petition requests. These relationships persisted for

the other variables even after offense seriousness was introduced as

a control variable. In the other city, age, sex, family stability,

year, arresting officer, and prior arrest record affected petition

requests. This analysis suggested that not only did characteristics

of the person and legal case characteristics have an effect on police

dispositions, but that characteristics of the police also influenced

the intake decision.

One of the most thorough of the studies of intake recommendations

was conducted by Terry (1967). He was interested in uncovering the

criteria used by probation officers in making intake decisions. He

analyzed official records utilizing multivariate contingency table

analytic techniques, computing the rank-order coefficients. First,

Terry (1967: 177) noted that 30 per cent of the youths who reached the

intake level had their cases dismissed. For those juveniles who were

referred to court, the seriousness of the offense, the number of prior

offenses, and the age of the juvenile were the only factors found to

be significantly associated with intake recommendations.

Ferdinand and Luchterhand (1970) examined police arrest records

for juveniles and then combined them with attitude measures obtained

from the juveniles. They found that the race of the offender had a

definite effect on police dispositions. Blacks were more likely to

be referred to court than were whites. This was true even when type

of offense, age, and sex, were controlled. Ferdinand and Luchterhand

(1970) also found that black arrestees scored higher than whites on

the attitude measure "Authority Rejection." The authors suggested

this attitude may have been apparent to investigating officers through

the demeanor of the black youths. Because deference to authority was

important to the officers, differences in demeanor may account for ob-

served race differences in dispositions.

Weiner and Willie (1971) examined the effects of race and SES on

police juvenile officers' decisions to either dismiss a juvenile's

case, send the case to a social service agency, or to send the case to

court. They found, in both cities studied, there were no significant

race or SES differences in court referral rates. Blacks, however,

were overrepresented in the initial arrest figures, as were lower-SES

youths. Weiner and Willie introduced no controls for other variables,

nor did they examine patterns in legal variables that might enter into

decision processes. The authors concluded that about equal proportions

of black and white youths, and lower- and upper-SES youths, were re-

ferred to court by juvenile intake officers. They argued that any pat-

terns of differential treatment by race or SES were to be found at the

police contact level, rather than at subsequent intake decisions.

Arnold (1971) was interested in the effects of legal and nonlegal

factors on intake decisions made by probation officers. Using official

records and tabular analytic techniques, he examined the relationship

between race/ethnicity and intake decisions. He found that minority

group members were more likely than majority group members to be re-

ferred to court, and that blacks were referred to court even more often

than Hispanics. Controls for marital status of the juvenile's parents

did not reduce this relationship between race/ethnicity and court re-

ferral. However, when offense seriousness, number and seriousness of

prior and concurrent offenses, and the delinquency rate of the juven-

ile's neighborhood were controlled separately, there was a substantial

moderation of the relationship though it was not completely eliminated.

Arnold concluded that intake officers did not necessarily favor the

majority group members. Instead, they applied the law with "excessive

severity" to members of minority groups.

Thornberry (1973) examined the effects of prior record, offense

seriousness, race,and SES on intake dispositions. He found that as

offense seriousness increased, so did the likelihood of petition recom-

mendation. Likewise, youths with prior records were more likely to be

petitioned to court than first-time offenders. Blacks were found to

receive petitions more often than whites, and lower-SES youths were

also more likely to be recommended for a petition than were those of

higher social status. These associations between the sociodemographic

variables and intake dispositions remained statistically significant

even after controls were introduced for offense seriousness and prior

record. Thornberry's data provided strong evidence of differential

treatment of juveniles based on nonlegal considerations.

Thomas and Sieverdes (1975) sought to determine the extent to

which social factors altered the degree of association between legal

case characteristics and case dispositions at the juvenile court

intake stage. Using official records and rank-order coefficients,

they examined the effects of offense seriousness and prior delin-

quency on case dispositions. In a simple rank-ordering of the ef-

fects of the independent variables, they found that the most impor-

tant factor was offense seriousness, followed in order of importance

by age at first offense, age at present offense, number of codefen-

dants, race, family stability, and sex. Social class of the juvenile

and number of prior offenses were not related to disposition at the

zero-order level. While several social factors were associated with

disposition, Thomas and Sieverdes pointed out that it was offense

seriousness, a legal variable, that had the greatest impact on the

disposition decisions of the intake officers.

To determine whether the effect of offense seriousness would main-

tain its impact in a multivariate analysis, controls were introduced

for sociodemographic variables. The relationship between offense

seriousness and disposition did vary across values of sex, prior

record, race, social class, family stability, number of codefendants,

and age at first and most recent offense. There were similar results

for prior legal record, although the relationships generally were not

as strong. The authors argued that the social meaning of the alleged

offense was quite different to the intake workers depending on the

particular social category into which the juvenile fit.

Cohen and Kluegel (1979a) investigated the decisions of probation

intake officers in two separate jurisdictions to determine the impact

of legal and nonlegal variables on case dispositions. They analyzed

official court records through the use of exploratory log-linear anal-

ysis. The results of their analysis showed that sex of the child,

court philosophy (due process versus therapeutic), and the type of

offense proved to be most strongly related to case dispositions. Race,

parental income, and family stability had no significant direct or

indirect effects on case dispositions. Juveniles with prior records

were more likely to be referred to court, and "idle" youths with no

prior record were also referred more often than were youths occupied

in conventional activities. Thus, while race and social class did not

appear to be related to intake decisions, other nonlegal factors were


The literature on intake processing does not exhibit the same

degree of consistency with regard to the roles of sociodemographic and

other nonlegal factors shown in the police-juvenile contact studies.

However, at least one "legally irrelevant" variable appears as a deter-

minant of intake decisions in all of the studies where more than race

and SES are considered (i.e., Weiner and Willie, 1971). In the cases

where race and/or SES differences were noted, they were always to the

disadvantage of minority group and lower-SES youths. On balance, how-

ever, evidence of discrimination against youths from less powerful

social groups remains problematic.

The inconsistency among the studies of intake processing in terms

of the effects of traditional theoretically important sociodemographic

variables does not, at this point, allow generalizable conclusions.

The role of the methodological techniques employed must also be con-

sidered, since these studies vary considerably in terms of technical

sophistication. The differences in methodologies and the inconsist-

ency of the results among these studies of intake processing are also

characteristic of the research conducted at other stages of the ju-

venile justice process.

Detention decisions

Sometimes intake workers must also decide whether or not to detain

an arrested youth while other decisions are made about his or her case.

Detention consists of holding the youth in a secure facility until a

judicial disposition can be obtained in the case. Euphemisms aside,

detention is the jailing of a juvenile awaiting subsequent hearings or

final disposition. The National Council on Crime and Delinquency

(cited in Cohen and Kluegel, 1979b), has provided a list of criteria

for deciding when a youth should be detained. They suggest that a

juvenile be held in detention when 1) it appears likely that he or she

will commit an offense dangerous to self or the community before court

disposition can occur; 2) there is a danger of running away; or 3) the

juvenile must be held for adjudication in another jurisdiction. Ideal-

ly, the decision to detain a youth should be based on the criteria and,

as with other decision points, there should be little evidence of dif-

ferential treatment attributable to sociodemographic variables.

Very little empirical research has been reported on this stage of

the juvenile justice system (Cohen and Kluegel, 1979b). Available

information suggests that legal factors are important components of

decisions to detain a juvenile (Sumner, 1970; Pawlak, 1977; Cohen and

Kluegel, 1979b). Sumner (1970, cited in Cohen and Kluegel, 1979b),

found that prior history of running away and prior detention experi-

ence were related to detention decisions. Pawlak (1977) observed that

juvenile code offenders were more likely to be detained than were

criminal offenders. Cohen and Kluegel (1979b) found that in addition

to the effect of prior legal history, the philosophy of the court in

which the juvenile was processed (i.e., therapeutic versus due process)

was associated with the decision to detain a juvenile. Therapeutically

oriented courts tended to use detention more often than due-process ori-

ented courts.

Sociodemographic and other nonlegal factors have shown inconsistent

effects across studies of detention decisions. For example, Sumner

(1970, cited in Cohen and Kluegel, 1979b) found no nonlegal variables

related to detention decisions. Pawlak (1977), on the other hand,

found that the sex of a youth was associated with the decision to de-

tain. His data also showed an interesting relationship between the

use of detention and the availability of detention facilities. Juris-

dictions with detention facilities detained youth more often than

those without. He also noted that whites tended to be detained more

often than nonwhites. Unfortunately, no control variables were used

in his analysis.

Cohen and Kluegel (1979b) found two nonlegal factors to be associ-

ated with detention decisions, net of prior record and court philosophy.

Like Pawlak (1977), they noted a tendency for females charged with

status-type ("decorum") offenses to be detained more often than males

charged with similar offenses (Males tended to be detained more often

than females when both were charged with more serious offenses). The

"present activity" (i.e., unemployed, dropout versus employed, in-

school) of the juvenile also had a significant impact on the likeli-

hood of detention. Juveniles who were not employed or in school were

most likely to be detained.

General conclusions about patterns in detention decisions are

difficult to make because of methodological inconsistencies in much of

the research. Pawlak's (1977) analysis, for example, is limited to

bivariate relationships, and this weakens the interpretation of his

results considerably. The review of Sumner's (1970) work by Cohen

and Kluegel (1979b) suggests that she employed multivariate tabular

analysis. This made it difficult to assess the strength of zero-order

relationships between decisions and nonlegal variables. Cohen and

Kluegel's (1979b) is the only study of detention that examines both

bivariate and multivariate relationships. Their analysis indicated

that nonlegal factors play an important role in detention decisions

even when legal variables are used as controls. At present, it seems

that legally irrelevant factors do enter into detention decisions.

However, since this conclusion is based on only a few studies, further

research is needed before any generalization about detention can be

made with confidence.

Juvenile court decisions

Decisions made at the court level are among the most critical in

the entire juvenile justice system. Judges decide whether a case may

be dismissed, whether some form of community-based treatment or proba-

tion may be required of the defendant, or whether the juvenile may be

committed to a residential treatment facility (e.g., a "training

school"). Thus, it is little wonder that judicial decisions have been

a frequent research topic for sociologists studying the juvenile jus-

tice process.

One of the earliest studies of juvenile court decision-making was

conducted by Terry (1967), who found that both legal and nonlegal fac-

tors influenced juvenile court dispositions. Terry (1967: 180) con-

cluded that the court personnel utilized a broad range of factors in

making decisions. These tended to be based partially on legal con-

siderations, but decision-makers were also influenced by the situation

in which the offense occurred and by the biography of the offender.

The number of prior offenses the juvenile had committed was the most

important consideration used by judges. Prior record had no relation-

ship to disposition, however, when the offender's age was controlled.

The number of persons involved in the offense, and the involvement of

adults in the offense were also found to be significantly related to

the disposition received. Important nonlegal factors included the age

of the offender, the sex of the offender, and the degree of commercial/

industrial development of the neighborhood where the offense occurred.

These did not diminish significantly with the introduction of control

variables. Terry (1967) did not introduce simultaneous controls for

more than one variable at a time.

In a similar study, Arnold (1971) examined the effects of several

legal and nonlegal factors on juvenile court dispositions. At the

bivariate level of analysis, minority youth were more likely to be com-

mitted to the state juvenile authority than were majority youth. Par-

ents' marital status and the delinquency rate of the neighborhood were

also found to be significantly related to judicial decisions. Children

from broken homes were more likely to be sent to training school than

were children from intact families. Youth from neighborhoods with high

delinquency rates were most likely to be referred to the custody of the

youth authority, as were those with more serious offenses and prior

records. Youths charged with more serious offenses and those with

prior legal records also had high rates of commitment. When simul-

taneous controls for all other variables were introduced, the effect

of minority group status did not disappear. That is, with all other

factors considered, the race of the juvenile still had a significant

impact on the type of disposition received. Arnold (1971: 223) noted

that blacks received harsher treatment than did Hispanics, and that

both groups were more harshly treated than white youth.

Ferdinand and Luchterhand (1970) were interested primarily in the

effect that race had on juvenile court dispositions. They found that

blacks tended to receive harsher sentences than white youths. However,

when controls for family status, prior record, and offense seriousness

were introduced, there were no significant differences between whites

and blacks. On this basis the authors concluded that race did not

have a significant impact on judicial decisions when other "relevant"

factors were taken into consideration.

A study conducted by Scarpitti and Stephenson (1971) compared the

factors that discriminated among adjudicated delinquent boys who

received probation, who were assigned to either a nonresidential or

residential treatment program, and those who were committed to a re-

formatory. Utilizing percentage comparisons among the groups, they

found that there was a slightly greater tendency for those assigned

to the reformatory group to be from lower-SES families, to be school

drop-outs, unemployed, and black than those assigned to the other

treatments. The reformatory boys more often than those in the other

groups also tended to be charged with serious offenses and to have

prior legal records. The authors then controlled for prior record,

offense type, and age at first court appearance by using a "delin-

quency risk" scale. Under controlled conditions, they found that

blacks were not treated disproportionately more harshly than were

whites. They argued that it was this "delinquency risk" factor that

judges used in making their decisions about what disposition to ren-

der, rather than any single legal or nonlegal factor.

Utilizing the Philadelphia birth cohort data collected by Wolf-

gang and associates, Thornberry (1973) explored the relationships

between race, SES, and sentencing in the juvenile court. At the bi-

variate level, the severity of sentence was associated with race, SES,

offense seriousness, and prior legal record. Blacks, lower-SES youths,

those charged with serious offenses, and those with prior legal records

were accorded the harshest sentences. Controlling for offense serious-

ness and prior record, Thornberry observed that both race and SES ef-

fects remained. He used this as evidence of differential processing

in the juvenile justice system.

Thomas and Cage (1977, see also Thomas and Fitch, 1975) examined

official court records to determine the associations among a variety of

"legally irrelevant" variables and the severity of juvenile court dis-

positions. In a rank-ordering of correlational analysis results, they

found that the following were significantly related to dispositions:

1) prior record; 2) offense type; 3) school status; 4) complainant;

5) judge; 6) sex; 7) race; 8) SES; and 9) home situation. When con-

trols for offense seriousness were introduced some of these relation-

ships were found to be even stronger. It was noted that males were

more likely than females to receive harsh dispositions, blacks received

harsher treatment than whites, school drop-outs were accorded more

severe sanctions than those enrolled, and youths from broken homes

were treated more harshly than those from intact families. Holding

constant the effects of prior record, similar, although weaker results

were obtained for race and school status. Controlling for prior

record, they also found that citizen-initiated complaints (as opposed

to police- or school-initiated) produced the least severe dispositions.

On this basis they argued that the severity of sanctions was partly

determined by who the youth was, who reacted to their behavior, and

the judge who heard the case.

Cohen and Kluegel (1978) explored the use of legal and nonlegal

factors in two juvenile court jurisdictions and found no relationship

between dispositions and either race or SES. They did, however, find

that "present activity" was a correlate of sanction severity. Only

legal factors were found to be good explanatory variables in their

long-linear analysis, and the legal variables offense seriousness and

prior record provided the majority of the explanatory power. However,

they did find "present activity" to be a correlate of sanction severity

also. As a result, they concluded that offense seriousness and prior

legal record were the best predictors of juvenile court dispositions.

Carter and Clelland (1979) utilized ordinary least squares regres-

sion techniques to study the effects of several legal and nonlegal vari-

ables on juvenile court dispositions. They divided offenses into "tra-

ditional" (i.e., criminal) and "status/victimless" crimes, following a

Neo-Marxist formulation. Severity of disposition for traditional

crimes was predicted by prior court referrals, social class, and pre-

sence of an attorney. That is, net of other variables, youths who had

prior records in court, were from lower-class backgrounds, or retained

an attorney were more harshly treated. In the case of "status/victim-

less" crimes, significant net effects were found for prior court record,

social class, and number of current charges. Juveniles with prior court

contacts, from lower-class backgrounds, or who were charged with multi-

ple offenses received harsher dispositions. Carter and Clelland (1979)

argued that the effect of social class would be greatest for the "status/

victimless" crimes, and their hypothesis was supported. The effects of

SES were consistent throughout, but other nonlegal factors washed out

in the multivariate analyses. Overall, legal case characteristics pro-

vided the best predictive power in their analyses.

One of the more recent studies of juvenile court decision-making

was conducted by Horowitz and Wasserman (1980). Using ordinary least

squares regression techniques, they examined the effects of legal and

sociodemographic variables on sentence severity. They found that the

best predictor of sentence severity, net of other variables, was wheth-

er or not the child was having problems at home. Juveniles who were

having problems were more likely than those living in harmony with

parents to receive a harsh sanction. The second strongest predictor

was previous arrest record. Youths with prior arrest records received

more severe sanctions than those with no prior arrests. The third

best predictor was school behavior. Children who were having behavior

problems in school received harsher sanctions than those having few or

no problems. The final significant predictor of sentence severity was

offense seriousness. The more severe the offense, the harsher the sen-

tence. Horowitz and Wasserman (1980) noted that there were no signif-

icant intercorrelations among these nonlegal factors and other socio-

demographic variables (e.g., age, race, and sex) that might indirectly

lead to differentially severe sanctioning. They concluded that certain

nonlegal factors should be expected to impact significantly on juvenile

court dispositions, since the juvenile court's foundation in discretion-

ary treatment was grounded in the use of such factors.

The preceding review of the empirical literature on juvenile court

decision-making shows that legal factors are uniformly associated with

dispositions in all of the studies. Generally, they are the most impor-

tant considerations employed by judges in deciding how to handle a de-

linquency case. The nature and extent of the impact of sociodemographic

and other nonlegal variables is not as clear. Cross-study variation in

the associations among nonlegal factors and dispositions does not allow

firm assessments of their strength and contributions to juvenile court

decision-making. However, with one exception (Cohen and Kluegel, 1978),

at least one legally irrelevant variables appears as a significant

predictor in all the studies reviewed here. This suggests that non-

legal factors play a role in judicial dispositions, although further

research is needed to specify their importance in judicial decision



Taken alone, the number of studies discussed in this review of

literature on differential processing should convey that this is a

topic that is interesting to sociologists. Unfortunately, the sum of

results from these studies is so ambiguous that a definite general

statement is extremely difficult. It appears, overall, that nonlegal

factors do play a role in juvenile justice decision-making. Determin-

ing how much of a role, and whether their use constitutes discrimina-

tory practice is not possible at present. This is due, in part, to

the fact that there are very few studies of juvenile justice decision-

making that employ sophisticated multivariate analysis techniques.

This criticism is especially germane when studies of intake disposi-

tions are examined. In the last section of this chapter the studies

of intake decision-making outlined earlier will be reexamined, this

time paying closer attention to the conceptual and methodological is-

sues raised in the review essays. Doing this should help to establish

the context against which the present study should be evaluated.

The fact that prior researchers have found nonlegal factors to

be associated with intake decisions has been pointed out in the pre-

vious review section. It was noted there that there is little consis-

tency across studies in terms of just what nonlegal factors have an

effect. One possible reason for this is the different methodological

approaches employed in these studies. The level of analytic sophisti-

cation (bivariate or multivariate) employed in the intake studies is

presented in Figure 1, along with other summary characteristics of the

previous intake research.

Figure 1 presents in summary form an indication of the variation

in the analytic techniques used across studies. Three studies (Arnold,

1971; Thornberry, 1973; Cohen and Kluegel, 1979a) have used multivari-

ate techniques to control simultaneously for the effects of several

independent variables. In all three studies, nonlegal factors were

found to be related to intake dispositions, though the same nonlegal

factors were not measured in all three studies. Likewise, in the

studies at the bivariate level of association, and in those studies

where controls were introduced for only one or two variables at the

same time, nonlegal factors with few exceptions (Weiner and Willie,

1971) have been found related to intake decisions. The effects of

race and SES, often the critical variables of interest in these stud-

ies, have not been found to be statistically significant above the

partial correlational analysis level.

Second, the sources of the data for these studies do not appear

to account for differences in their findings. All of these data, with

the exception of attitudinal data (Ferdinand and Luchterhand, 1970),

were taken from official police or court records. Of course, differ-

ences may be attributable to jurisdictional variation in the type of

information recorded on official documents. This is especially true

of information concerning parents' marital status, school-related

NOI93 I u ~: > 3 L: uj l )

31VIXVAInnI1 x xlx x
SIVIIVd xx x xx x x
31VIbVAI8 x x x x x x x
NOIIV301 P 0-. 1, a- u ul u u

31Vb IN3nbNNI130
Q3AlOANI Sb3HIO o -o
aMo3OR bOIbd ual x -o o .
3SN33JO o u-a x .0 u

aOOHa08H9I3N ~
SS3N371OI i u
A1IWVJ .,- x ,' x
X3S x o .- x u
39V a0 .0 x .0
S3S tc x x u .0 x
33V 'o m x -. x uu x


S o_
00 r- r--
.1--~, -r- -- o -

10 a) *a) ..0 an )-

5.- LLi U M- 0 0 0 cu =-
o s C) U U- o = CE- 0- a
C5 F- X L- _1 3r. m <: u i:4



3 C
,_- 0
i ,- 4-

S0 U
(U 0'-1

04-) CC
4) 4.) 4 )
*i- *-*- *I -

4 *- -I a.)- .
" .0 C 0 .Cl
0 4> .r- *-) C 4-C
0 3 *.- C.
.4. C C C U)
r 000 U 4-.
4-31 4. 4.) r- = In
C *- *- 0 0 J 0
0 In tI U0.U C >-
* 0 0 0
*r- -- *,- -r- a. *- Z
In 4-) 41 4.1

0.---,--.- .Q
(0C CC C S. 4-


U) In on oI 0 0.-i- U)
E I0 (A0 0 I II I I

x M.0 uw L u'i 3

information, and SES. Even when the same information is recorded in

all places, the measures used may vary from time to time, from office

to office, and from official recorder to official recorder.

Third, researchers do not always consider all the variables that

are theoretically relevant. This makes it difficult to standardize

results across previous research. One step that might help in the

future is for all researchers to include at least the major and most

often accessible information in their analyses, adding other variables

in subsequent models as they are available. This would allow examin-

ation of the effects of the standard theoretically important indepen-

dent variables on dispositions, and it would allow as well for the

addition of interesting information that is not always available.

Fourth, studies of "intake" are not necessarily dealing with the

same thing. Intake can operate at two different points in the juven-

ile justice system. In some jurisdictions intake is carried out at

the police level, while in others it is accomplished by another agency,

usually a probation department under the control of the court. In

four studies of intake decisions reviewed above (Goldman, 1969; Ferdi-

nand and Luchterhand, 1970; Weiner and Willie, 1971; McEachern and

Bauzer, 1967), the police department handled intake proceedings. In

the remainder of the studies (Terry, 1967; Arnold, 1971; Thornberry,

1973; Thomas and Sieverdes, 1975; Cohen and Kluegel, 1979a), intake

was handled by officers of the court. Intake by the police represents

a different degree of juvenile justice system penetration than when a

court-appointed probation officer serves the function. Further, the

philosophy of law enforcement may impact differently on intake

decision-making than does the emphasis on and philosophy of rehabil-

itation which is more characteristic of probation workers.

Fifth, studies of intake vary in terms of the geographical re-

gion from which the data were taken. The research reviewed here drew

samples from various parts of the country. Some data were taken from

the south (Arnold, 1971; Thomas and Sieverdes, 1975; Cohen and Kluegel,

1979a), some from the east and northeast (Goldman, 1969; Weiner and

Willie, 1971; Ferdinand and Luchterhand, 1970), others from the mid-

west and west (McEachern and Bauzer, 1967; Terry, 1967; Cohen and

Kluegel, 1979a). However, the absence of clear regional patterns

indicates that region of the country seems to have little to do

with the results obtained at this stage of the juvenile justice pro-


Sixth, studies done at different points in time may produce

differing results. If historical events have a major impact on the

character of juvenile justice, this may be-reflected in the extent

to which nonlegal factors are used in decision-making. With the ex-

ception of the Cohen and Kluegel (1979a) data, all of the studies

reviewed here examined delinquency cases processed prior to 1970.

Some of these data were taken from cases reaching back into the

1940's (McEachern and Bauzer, 1967). Unfortunately, it is difficult

to know whether time differences exist without longitudinal data col-

lected from the same jurisdiction. Ideally, a test of the time hypo-

thesis would require longitudinal data from several different juris-



The preceding review of the theoretical and empirical literature

has been designed to set the stage for the analysis of intake level

decision making. The two major foci of juvenile justice decision mak-

ing studies, as delineated by Horowitz and Wasserman (1980), are

shared by the present study. These are: 1) the determination of what

legal and nonlegal factors are related to intake recommendations; and

2) whether such factors are used in a manner that affects certain

social categories of youth, particularly the relatively powerless

segments of the youthful population, in an adverse manner.

The present study will examine intake decisions in one Florida

county. It will address three questions. First, what factors are

related to intake decisions? Second, is there differential treatment

of juveniles on the basis of these factors? Third, does the use of

any of these factors indicate discrimination against youths from rela-

tively powerless social groups? The data to be analyzed and the anal-

ytic techniques to be employed in exploring these questions will be

presented in Chapter Three.




Determining whether or not juveniles from the social categories

that conflict and societal reaction theorists define as powerless are

treated differently in the juvenile justice system requires empirical

research. In the present study official records from the agency that

performs the intake function in Florida will be analyzed in order to

address the issue of differential treatment. This chapter describes

the data source and collection of the data, as well as the way the

data were organized in order to explore this question. The statisti-

cal techniques that will be utilized to determine whether or not dif-

ferential treatment exists will be discussed in the final section of

the chapter, and the manner in which the results may be interpreted

is noted.

Data and Sample

Data for this study were provided by the Florida Department of

Health and Rehabilitative Services (HRS) Youth Services Program Office.

Florida's HRS is an umbrella state agency that includes most of the

state's social programs. These include social welfare, child, youth,

family, aging, health, and mental health programs. As well as serving

a general intake function, intake workers in HRS are responsible for

reviewing the cases of all juveniles arrested by the police, and for

making a recommendation on case disposition to the State Attorney.

Intake in Florida operates independently of the police and the court.

It is an intermediate step between arrest and the State Attorney's

decision about whether or not to prosecute a case.

The job of the intake workers in delinquency cases is to inter-

view juveniles charged with offenses that would be crimes if they

were adults. This must be done within 20 days, otherwise the intake

worker must recommend that a petition to court be filed. Intake

workers also may interview the arresting officer, the victim, and

the parents of the child. After reviewing the information gathered

through the interviews, the intake worker is required by law to make

a disposition recommendation to the State Attorney. The final deci-

sion on court referral is made by the State Attorney. To the extent

that the State Attorney's office is influenced by the intake worker's

recommendation, this recommendation carries considerable weight in

the juvenile's experience with the juvenile justice system.

The present analysis is based on data drawn from records gener-

ated by HRS intake counselors in one county in north Florida (Alachua).

Single-intake records for the time period July 1, 1977,to September 30,

1979, were obtained, yielding 1,507 delinquency cases. Of these 1,507

cases, 1,404 had information on the variables used in this study.1

Alachua County was chosen as the focus of this study for several

reasons. First, this county provides a fairly heterogeneous popula-

tion along the dimensions of socioeconomic status (SES), region of

origin, race/ethnicity, educational attainment, and religious affilia-

tion. Second, the author was familiar with both the county intake

office and the State Attorney's office and had conducted lengthy inter-

views with staff members who provided information about their decision-

making processes. Their accounts offered much valuable information

that allowed the author to base many of his research decisions on in-

sight rather than speculation. The two offices showed a high degree

of continuity in both philosophy and practice during the time period

under study. Third, Alachua County may be considered fairly represent-

ative of other Florida counties. Patterns in intake decisions do not

appear to differ markedly from those found in nearby Florida counties.

In fact, an analysis of the larger data set from which these data were

drawn shows little variation across seven counties (Potter and Frazier,


The data to be analyzed here come from Intake Data Cards. These

cards contain a wide variety of information. A copy of each form is

reproduced in the Appendix. Several types of information are collected

on the Intake Data Card. Much of the sociodemographic and incident-

related information is recorded at or shortly after the intake inter-

view. The remaining information, regarding decisions at subsequent

processing, is recorded at later dates.

The information gathered on these forms focuses on the child,

the family, the offense and situation, and the outcome at various

decision points. As Cicourel (1968) has noted, official documents

never contain all the background information an analyst might wish to

have. Intake Data Cards are no exception. They are limited summaries

of the juvenile's case file and seldom contain information about vari-

ables such as family income, parents' occupations and/or their educa-

tion. Instead, they focus on the basic sociodemographic and legal

characteristics of individual juveniles and their cases.

The Intake Data Cards

During the time period that this information was collected, two

forms for the Intake Data Card were used by HRS personnel (see Appen-

dix). Most of the information was recorded in the same manner on each

form. However, some information categories were dropped from the

second form and some other information was collected in expanded or

altered form. For the purposes of this study, the information has

been divided into three categories. The first category included all

information about the sociodemographic characteristics of the juvenile,

the second category included the legal characteristics of the case and

the juvenile, and the final category included information about process-

ing and prior agency experience of the individual.

Section A on both forms contained sociodemographic information

about the client. Client's name, date of birth, sex, race, Title XX

eligibility, and the county of residence were all recorded in this sec-

tion. Recording of this information remained unchanged from one form

to the other.

The legal characteristics of each case were found in different

sections on the two forms. On the first form (Form A), items dealing

with the reason the juvenile was referred to intake (offense) and the

legal status of the client at the time of referral were recorded under

items G and H. This same information was found in items E and F of

the second form (Form B), and the recording formats were unchanged

from one form to the next. The number of charges against the client

at the time of arrest (number of log entries) was found in item S of

Form A and item P of Form B. Again, the recording format remained

unchanged in the two forms.

The client's case status with HRS (new versus current case) was

found in item T of Form A and item Q of Form B. Information about

the disposition of the case was recorded in a different manner on

each form. On Form A, the intake worker's recommendation to the

State Attorney was recorded under item L. Item M contained informa-

tion about the intake worker's recommendation to waive the case to

adult court, as well as the State Attorney's action on this matter.

On Form B, the intake worker's recommendation to the State Attorney

was recorded in item J. It was expanded to include options ranging

from recommending that no petition be filed, to recommending that the

client be indicted by a Grand Jury. Item K contained the State At-

torney's action concerning the case, with options ranging from no

petition filed to the filing of an indictment.

Data about case processing history were recorded on each form.

Both forms contained dates showing when the client was brought to in-

take for the first time, when the case was received, and the date the

case was closed. Data pertaining to the final disposition of the case

were recorded in different manners on the two forms. Form A provided

spaces for recording the court decision, the judicial decision, the

nonjudicial decision, and the use of and the amount of restitution.

Form B had spaces for most, but not all, of this same information on

case disposition. However, it was broken down into two slightly dif-

ferent categories: nonjudicial and judicial actions. Specific types

of sanctions and interventions that could be used in the case dispo-

sitions could also be recorded with a yes or no.

Some of the variables used in the following analysis did not ap-

pear on the Intake Data Card in the form that they appear in the anal-

ysis. In order to merge the two formats into one data set, it was

necessary to recode some of the legal and processing variables in

both formats. Sociodemographic variables did not require recoding.

While in some cases, data categories had to be sacrificed in order

to provide continuity between the two formats, the final data set

retains the overall integrity of the original information.

Measures, Variables, and Codings

Seven independent variables and three dependent variables have

been constructed for the purposes of this study. Independent variables

include the sociodemographic variables age, sex, and race; as well as

the legal variables offense seriousness, prior legal record, number

of charges, and agency case status. The variable names, codings, and'

distributions of these variables are presented in Table 3.1.

The major dependent variable to be examined in the present study

is the intake worker's recommendation to the State Attorney concerning

the disposition of a juvenile's case. Intake worker's recommendations

were recorded differently on the two recording formats (see Appendix A).

On both forms the intake worker could recommend that the State Attorney

Table 3.1: Variable Names, Codin~gs, and Distributions

Variable Name



Juvenile's Age

Juvenile's Sex

Juvenile's Race

Agency Case Status

Prior Legal Record

Offense Seriousness

Number of Charges

Intake Recommendation

0 = 13

1 = 13-15

2 = 16-16

0 = females

1 = males

0 = whites

1 = blacks

0 = new

1 = current

0 = no prior record

1 = prior record

0 = misdemeanor

1 = felony

0 = one

1 = two or more

0 = no petition

1 = petition









Table 3.1: Variable

Names, Codings, and Distributions

Table 3.1-Continued.

State Attorney's Action

Judicial Decision

0 = no petition filed

1 = petition filed

0 = dismissed

1 = adjudication withheld

2 = commitment



not file a petition to court, or that a petition to court be filed.

On Form B, intake workers could also recommend that the State Attorney

waive the juvenile to adult court, or that an indictment be filed by

a Grand Jury. These latter recommendations were made so seldom (in

less than 1 percent of the cases), that they were included in the

"file petition" category. This allowed for easy merging of these

information categories on the two intake data forms. Any case that

received a "no petition" recommendation was assigned a value of zero

(0). Cases that received a recommendation for some form of action

beyond the prosecutorial level were interpreted as "file petition"

recommendations and were assigned a value of one (1).

The second dependent variable that will be examined in this study

is the State Attorney's decision on whether or not to file a petition

to court in a particular case. State Attorney's action was recorded

differently on the two recording formats utilized, much like the

Intake Recommendation (see the Appendix). In order to reconcile the

two recording formats, and because some options were used so infre-

quently, State Attorney's action was dichotomized. Any case that was

not filed on was assigned value of zero (0). Cases that were filed on

or waived to adult court were assigned a value of one (1).

The third dependent variable to be examined in the present study

is the judicial disposition. Unlike the two previous dependent vari-

ables, there were many categories of judicial disposition that could

be recorded on both recording formats (see the Appendix). There were

reduced to three categories. Each of these levels of judicial

disposition represents a different depth of penetration into the judi-

cial process. In addition, each category carries with it a differing

degree of potential sanction.

Cases that were handled nonjudicially or dismissed received the

least severe sanction. Cases with adjudications withheld constitute a

greater degree of sanction. Those cases that were adjudicated and

placed in some form of commitment represent the strongest sanction

applied to juveniles, in terms of both length of probation or confine-

ment, and the qualitative dimensions of the controls imposed. These

three levels of judicial disposition are viewed as forming a scale

with more or less equal intervals between the levels, such that the

distance between nonjudicial actions or dismissals and adjudication

withheld is the same as that between adjudication withheld and adjudi-

cation and commitment.

Following this logic, cases that were termined prior to a judi-

cial disposition or that were dismissed by the judge were included in

the category coded zero (0). Cases in which adjudications were with-

held were assigned a value of one (1). Finally, cases that received

some form of commitment to HRS or detention were assigned a value of

two (2).

Of the seven independent variables used in this study, six were

dichotomous and one was categorized into three levels. Age at time of

referral to HRS was coded as a three-value variable for the purposes

of this analysis. Originally it was created as a continuous variable

by subtracting the number of days between the date of the client's

birth from the date of referral to HRS. The three categories were

created for two reasons. First, the three categories correspond closely

to those employed in daily decision-making by intake workers interviewed

by the author. The intake workers divided juveniles into groups on the

basis of age, with the perception that the juvenile's level of responsi-

bility rose as age increased. The intake workers did not see levels of

responsibility increasing on a year to year basis. Rather, they saw

levels of responsibility, with cut-off points at the ages noted below.

These levels are treated, as in the case of judicial disposition, as

equally spaced. Thus, the increase in responsibility attributed to a

juvenile was roughly the same at the 12 to 13 age point as it was at

the 15 to 16 age point.

Children below the age of 13 were perceived as deserving the least

severe sanctions of all juveniles. Youths between the ages of 13 and

16 were generally treated more harshly than younger children, but not

as severely as those above the age of 16. Juveniles 16 or older were

perceived as the group most responsible for their behaviors, and thus

generally subject to more severe sanctions. As the youths passed from

one age category to the next, then, the severity of the sanctions that

the intake workers felt were appropriate tended to increase. In the

criminological literature, evidence exists that juveniles age 16 and

older are among the segment of the population most subject to crimin-

alization (Turk, 1969; Quinney, 1970).

On the basis of these considerations, it was decided that these

three age groups constituted a scale. Therefore, children age 12 and

younger were included in the category coded zero (0). Juveniles aged

13 through 15 were included in the category coded one (1). Finally,

juveniles age 16 to 18 years were placed in the category coded two (2).

The mean age of the juveniles referred to HRS intake in Alachua County

during this time period was 15.2 years (sd = 2.1 years).

The remaining independent variables are dichotomous, and dummy-

coded, so that the zero values represent the category of the variable

that the literature suggests is least likely to be associated with

severe sanction. The category coded one includes those values of the

variable generally considered the most likely to promote severe treat-

ment. For example, on the variable sex of the client, females are

assigned a value of zero, and males a value of one. Race of the cli-

ent is coded such that whites are given a value of zero and blacks a

value of one.

HRS case status is a measure of the agency's familiarity with a

client. Clients who were not currently HRS clients were assigned a

value of zero. Juveniles who were current HRS clients, either for a

previous delinquency referral or for an abuse/neglect referral, were

assigned a value of one.

Prior legal record measures all prior delinquency-related con-

tacts with legal agencies. A value of zero was assigned to cases

with no prior legal record. Juveniles who had experienced some form

of prior legal processing, whether an "information only" at the police

level, or some action carried further through the juvenile justice

system, were assigned a value of one on this variable.

There were 39 offenses with which the juveniles could have been

charged on the two recording formats. These offenses were broken down

into two categories for this analysis. Misdemeanor offenses were as-

signed a value of zero, and felony offenses were assigned a value of

one. This dichotomy seems appropriate for two reasons. First, the

vast majority of the felonies were felonies of the third degree.

Since the maximum penalty for a third degree felony in adult court is

five years in prison, intake workers and State Attorneys tended to

equate these offenses in degree of seriousness. Second, while mis-

demeanors are divided into three degrees of seriousness in Florida

law, most offense categories were treated similarly by both intake

workers and State Attorneys (Frazier and Potter, 1982).

The number of possible charges in a case ranged from one to nine

on the Intake Data Card. Only 9 per cent of the cases were charged

with more than one offense, and almost none were charged with more

than two counts. As a result, juveniles charged with only one offense

were assigned a value of zero, and juveniles charged with more than

one offense were assigned a value of one.

Some important variables that have appeared in previous research

are missing from this analysis. These are primarily measures of SES,

family stability/unity, and school status variables. Variables that

are not often encountered in previous research, agency case status

and number of charges, have been included. Agency case status is

added because it allows an examination of the effect of being known

to the intake unit on the type of recommendation forwarded to the

State Attorney. This may give an indication of the effect that

bureaucratic knowledge of the case has on the type of decisions made

concerning the case in a quasi-legal setting. The number of charges

is included as an indicator of the juvenile's possible involvement in

delinquent activities.

Official records are always vulnerable to certain problems, and

these data are no different (e.g., Kitsuse and Cicourel, 1963; Cicourel,

1968; Needleman, 1981). One possible problem that exists is that,

within the intake office itself, there is no single, well-delineated

set of procedures for the recording of the information contained on

the Cards. New intake counselors are informally taught how to record

the information on an "as needed" basis. Given this, there is the

possibility that a degree of inconsistency might exist among the in-

take counselors on how and at what points they record the data. In

addition, the high personnel turnover characteristic of this office

introduces the possibility of inconsistent recording techniques among

the counselors. However, the rather basic nature of the information

recorded on the forms would seemingly help to aleviate these problems

in the present analysis.

The fact that the information recorded on the forms is obtained

at different points in time poses another threat to accuracy (e.g.,

State Attorney's action and later processing decisions and dates).

It is possible that intake counselors may be changed on a client's

case, so that some information received during the change over may be

missed. It is also possible that a counselor may simply miss receiv-

ing notification of decisions made at a later date concerning a juven-

ile's case.

A third concern arises because of the changing of recording for-

mats. Counselors, having gotten used to one form, may be confused by

the new form, particularly when information recording categories are

changed. That is, the counselors may not know quite how a certain bit

of information fits into a new scheme that either restricts or expands

the categories available to them. Thus, across the time span studied

here, some inconsistencies in recording certain types of information

might be expected. These data are vulnerable to this problem, although

the variables described above would seem to exhibit a high enough de-

gree of consistency to allay such fears in this analysis.

Overall, the data utilized in the present study do not seem to be

plagued by severe problems of the kind noted above. There are no dis-

proportionate amounts of missing data on any variable, and no patterns

in the missing data were noted. The problems inherent in official

records data may be lurking in these data, but to the extent that they

can be detected, they have not been found. In sum, the present data

are judged to be as sound as those sued in previous studies of intake

decision making. As with all such data, however, the results should be

considered in light of potential shortcomings. The data to be analyzed

here do seem to be sufficient to the task for which they are being used.

Analysis Techniques

The data were analyzed in three stages, beginning with cross-tabular

analysis and moving through bivariate correlation to multivariate logis-

tic regression techniques. A logistic regression technique is used be-

cause the dependent variables (intake recommendation and State Attorney's

action) are dichotomous. The actual values of the dependent variables

are zero (no petition) and one (petition). However, in a multivariate

analysis, the predicted probability of a petition recommendation may

range between these boundaries. Because these dependent variables are

dichotomous, relationships with the independent variables are likely to

be nonlinear, and the S-shape of the standard logistic distribution is

often appropriate. Logistic regression techniques model a nonlinear

relationship between the independent variables and the dependent vari-

able. They also constrain the predicted probabilities to fall between

the values of zero and one. In addition, logistic regression techniques

take into account the multiplicative nature of the relationships among

the independent variables.

While the values of the logit coefficient are unbounded, the prob-

abilities obtained are constrained to fall between zero and one (Hanu-

schek and Jackson, 1977; 188). The effect of each independent variable,

at any particular point on the logistic curve, can be obtained from the

logistic coefficient2. This transformation produces the predicted pro-

portional change value for the single independent variable, or the ef-

fect that a one-unit change in the value of that variable has on the

value of the dependent variable. In the analyses that follow, the pre-

dicted proportional change produced by each independent variable will

be examined. Given the theoretical framework guiding this research,

however, particular attention will be paid to the change in the prob-

ability of petition recommendation and petition filed produced by the

sociodemographic variables, since they are the theoretical indicators

of relative powerlessness.


IFamily income or some other measure of financial stability or
parental employment status would be beneficial for more fully testing
the relationship between SES and processing decisions, or between SES
and other variables that might have an impact on processing decisions.
However, the Intake Data Cards contain no information on variables
such as family income, parental employment status and the like. The
closest indicator of family income on the Intake Data Cards is Title
XX eligibility, a Federal family poverty-level indicator. Of the
1,404 clients processed through the intake office during the time
period studied here, only two were not considered eligible for some
Title XX funds on the grounds of family income. It seems unlikely
that these numbers accurately represent the true proportions of youth
above the poverty line who are referred to intake on delinquency
charges. There may very well be some selection bias in police arrest
probabilities, such that children from more affluent families are not
arrested as often as children from poorer families. It seems unlikely
that such a bias would be this extensive. Rather, it is more likely
that the intake workers do not fully understand the use of this pro-
gram, since such determinations of eligibility are more important at
a later stage in processing. For these reasons, Title XX was not used
as an indicator of family income in the present study.

2The change in the probability of a petition recommendation for a
one-unit change in the value of one independent variable, controlling
for the effects of the other independent variables, is B [P(1-P], where
B is the logistic regression coefficient and P is the probability of
being 1 at the point on the logistic curve that the effect of the
independent variable is evaluated (Hanuschek and Jackson, 1977).



This chapter presents the results of analyses of the data described

in Chapter Three. The analysis will begin at the bivariate level, with

these results being compared to those of previous studies that reported

bivariate measures of association. It then examines correlations among

the independent variables to determine whether there are any intercor-

relations that might present problems in a multivariate analysis. Log-

istic regression techniques are then used to determine the net effects

of the independent variables on intake recommendation. These results

will be evaluated in terms of prior multivariate research on intake

dispositions. In addition, the impact of intake recommendation and the

other independent variables on State Attorney's action and final judi-

cial disposition will be explored using multivariate analysis techniques.

Bivariate Analysis

The first step in the exploration of the effects of legal and

sociodemographic variables on intake recommendation involves determin-

ing whether or not these independent variables are associated with in-

take recommendation. This will be accomplished by examining the pro-

portion of youth receiving "no petition" versus "petition" recommenda-

tions for each value of the seven independent variables. These

proportions and raw frequencies are presented in Table 4.1, along with

the chi-square value for each association. Each of the associations

will be examined separately in the following discussion. The results

will be compared to those from previous studies of intake decision-mak-

ing that have presented bivariate analyses (Goldman, 1969; McEachern

and Bauzer, 1967; Terry, 1967; Ferdinand and Luchterhand, 1970; Weiner

and Willie, 1971; Arnold, 1971; Thornberry, 1973; Thomas and Sieverdes,


The age of a juvenile has been found to be associated with intake

decisions in several earlier studies (Goldman, 1969; McEachern and

Bauzer, 1967; Terry, 1967; Thomas and Sieverdes, 1975). In the pre-

sent study, there is also an association between the age of a youth

and the intake recommendation made in a case. Only 40 per cent of

the age group under 13 years old receive petition recommendations,

while the majority (62 per cent, respectively) of the youth in the

13-15 and 16-18 age categories are recommended for a petition to

court. It appears that being an older juvenile increases one's like-

lihood of petition recommendation.

Sex of a youth also has been shown to be associated with intake

dispositions in some previous research (Goldman, 1969; McEachern and

Bauzer, 1967; Terry, 1967; Thomas and Sieverdes, 1975), and some

authors expect it to be an important consideration in case processing

(e.g., Turk, 1969; Quinney, 1970; Chesney-Lind, 1977). An examination

of Table 4.1 shows that, while a majority of both males and females

(61.17 per cent and 51.60 per cent, respectively) receive petition

recommendations, the proportion of males who are recommended for

Table 4.1: Bivariate Associations with Intake Recommendation

No Petition
Percentage (N)

Percentage (N)


Age of Juvenile

Sex of Juvenile

Race of Juvenile

Agency Case Status

Prior Legal Record
No Priors

Offense Seriousness

Number of Charges
More than One

48.40 (121)
38.83 (443)

45.38 (354)
34.26 (209)

48.83 (481)
20.59 ( 84)

55.45 (412)
22.10 (139)

53.49 (399)
25.15 (163)

42.81 (530)
21.88 ( 35)

*All Chi-square values significant at

51.60 (129)
61.17 (698)

54.62 (426)
65.74 (401)

51.17 (504)
79.41 (324)

44.55 (331)
77.90 (490)

46.51 (347)
74.85 (485)

or below

.01 level.













Table 4.1: Bivariate

Associations with Intake Recommendation

petition to court is 10 percentage points higher than the proportion

of females receiving that recommendation. The chi-square test proves

to be statistically significant, indicating that sex differences such

as those found here are not easily attributable to chance factors.

Table 4.1 also shows that 65.74 per cent of the black youths re-

ceive petition recommendations for court while 54.62 per cent of the

white youths are recommended for a petition. This 11 percentage point

difference is statistically significant according to the chi-square

test. These results are consistent with some prior research (Goldman,

1969; Ferdinand and Luchterhand, 1970; Arnold, 1971; Thornberry, 1973;

Thomas and Sieverdes, 1975), as well as the theoretical literature

(Lemert, 1951; Turk, 1969; Quinney, 1970). Other empirical research,

however, has not found an association between race and intake disposi-

tions (McEachern and Bauzer, 1967; Terry, 1967; Weiner and Willie,


A juvenile's case status with HRS also proves to be associated

with the type of intake recommendation received. This variable may

indicate that a youth is currently on some form of probation for a

previous delinquent offense, or it may indicate that the youth is

under HRS supervision for an abuse/neglect or dependency referral. A

majority of all cases are given petition recommendations, but youths

who are already HRS clients are more likely to be recommended for pe-

tition than those who are new clients. There is a 28 percentage point

difference between these two categories.

In short, all of the independent variables are associated with

intake recommendation at the bivariate level of analysis. The distri-

bution of the legal variables suggests that the likelihood of petition

recommendation is greatest for children charged with felony offenses,

children with prior legal records, those charged with multiple offenses,

or children with prior contact with HRS. Similar results have been

noted by previous researchers with regards to the impact of legal vari-

ables on intake decision-making. The distribution of the sociodemo-

graphic variables suggest that blacks, males, or older juveniles are

more likely to be recommended for a petition to court than whites,

females, or younger adolescents. These bivariate associations are not

particularly convincing, however, since the introduction of control

variables may bring about changes in the patterns of the associations

noted above. Before conducting a multivariate analysis to explore

this possibility, a zero-order correlation matrix will be examined to

determine the nature of intercorrelations among the independent vari-


Pearson product-moment correlations (r) allow an observation of

the direction and strength of the relationships among the independent

variables. This discussion will focus on the apparent substantive

importance of each relationship. Thus, while coefficients achieving

statistical significance at or below the .01 level are noted in all

tables, statistical significance plays no role in the present analyses.

Rather, it is the magnitude and direction of these relationships that

are important. Along these lines, the first thing to note in Table

4.2 is the low intercorrelations among the independent variables.

< 0


o --"m
U- 0 C

0 O I
VO O e

0X *I C
<- o mn O

4']c 4 : -

0 L 00

C) i LO LO O

SC -
( a O o L LA C o L
S,-- L i] 0- ,- 0

0 L C) .-) .-
o L1-= F-. .- ay LA -_ LA

L i 0 0 ON -,- I
-" c .- -

4- C) o C) C C) C C

L* 0 a- U. f LA '.0

C 0 o 0 (Y 0

C-) O O C) Ln 0)
.0 C) 0 0 L- t o n t
I- I- 0 r- 9- 0 0 1- \j c..J i- o

C) r- C) C .
C >

LJ 0 LC %) CM- F4 r c -c'0 0
CD (Y C) a- 0 )- C<) C) C1 a-

0 +'> U0
V) < M -
o- .) ) C O ) OO ) t
C.C)Om1 rl m o Y

S- *a- a) a- i u U (-
a1 >., U) -__ 4. a) U)
C C U U C C a) C4 0 -
o i. U a) a) > a) i- S.. ) C
U- O Q 0 > > >(n L( U) a )

.0 WLJJ )J 4-laJ '4- oX >O i- h S ll-
(U E WI- UC 00^ 01 0-. Lii C)' Lt) a,- a4-
1. 4-IZ 'UC.? 0 WW XLJ UC. (UQ *-LiJ 4-Ll. O 0
'0 Ci-i t-'ct 3'-3 n3 (U ^0 0 CK 013: S- ; H40-3
;>| -i 00 -' '~3*-* --* fc0 > i- .-. *< d.'_* -- 4-)*

Eleven of the 21 correlations are below .10, and 10 of the correlations

range between .10 and .20. Only one coefficient is above the .30 level.

The correlations among the sociodemographic variables will be discussed

first, followed by a discussion of the correlations among the legal


Scanning Table 4.2 reveals only weak intercorrelations among the

sociodemographic variables. Examination of the correlations among the

sociodemographic and legal variables shows them to be somewhat stronger,

though still fairly weak. Among the legal variables themselves, some-

what stronger intercorrelations are noted, though again, they are still

fairly weak, with the exception of the correlation between agency case

status and prior legal record, which is fairly strong.

If this analysis were to end at the bivariate level, as in many

previous studies of intake decision-making, it might be argued that in

this sample blacks, males and older juveniles receive harsher treat-

ment than their counterparts. One could also argue that there is

differential treatment in intake recommendation decisions. However,

it would be necessary also to point out that legal factors have the

largest effects on intake recommendations. At the bivariate level,

youths with felony offenses, prior legal records, those charged with

more than one offense, or those who are current clients of HRS seem

more likely to receive a harsh recommendation than their counterparts.

Bivariate analyses, of course, do not allow a firm basis for draw-

ing either conclusion due to a lack of controls being introduced to

discover possible overlap among these independent variables. A

multivariate logistic regression was also performed in order to examine

the effects of each independent variable net of others that may also be

related to intake decisions.

Multivariate Analyses

Intake Recommendations

The results of a logistic regression of intake recommendation on

the legal and sociodemographic independent variables are presented in

Columns A, B, and C of Table 4.3.1 Of the 1,355 cases with intake

recommendation recorded, 146 had missing data on at least one of the

independent variables. As a result, a total of 1,209 cases were an-

alyzed.2 The predicted proportional change column (Column C) shows

the effect of each independent variable, evaluated at the mean value

of the dependent variable, net of the effects of the other independent

variables. The discussion of the coefficients will center on the sub-

stantive impact of each independent variable. For the purposes of this

discussion, any variable that alters the predicted proportional change

by .10 or more under controlled conditions will be considered substan-

tively related to intake recommendation.

An examination of Table 4.3 shows the legal variables to be the

best overall predictors of intake recommendation. Offense seriousness

shows the strongest effect on type of recommendation, with a felony

offense increasing the likelihood of a petition recommendation by

29.4 percentage points, net of other variables in the equation. Other

multivariate studies of intake decisions report similar findings

(Terry, 1967; Arnold, 1971; Thornberry, 1973; Cohen and Kluegel, 1979a).

c o Lfl ( 0 r- -n
o 0 O 0 rO

00 o i (3 r a)
o o 0 0 C-

* o 0 CJ .

ko 0 o C cc C

ko O O m C a
cV CD O mM r-. C CD O

0 Cn

a) a ) 0 0 C
CC <( CS *U- C o
.O L- i- r-. CD ) i- CM C0

a o r- S- cj a- S

)> > > Q 4-
SIn S- C,
01 0U r0 C/I Q 0 r1

0) r- L L)

0 ) 0 L 0 1 0 >) +

*r- *- *i S.- U

> > ) 0) 4C- E -
3 3 OT S- 5 3 C
'"a ^ '^ o- Z -i














Cohen and Kluegel (1979a) present results from a log-linear analysis

that control for the simultaneous effects of several independent vari-

ables, while the remaining researchers entered control variables one

at a time, offering only partial controls. Regardless of the various

levels of sophistication of the multivariate analytic techniques, how-

ever, offense seriousness has been found to be related to intake dis-

positions consistently in prior research.

Having a prior legal record produces almost as great a change in

a juvenile's probability of receiving a petition recommendation as

does offense seriousness. Youths with prior legal records are 27.3

percentage points more likely than those without prior records to be

recommended for a petition to court. Again, previous multivariate

analyses of intake decisions report similar findings (Terry, 1967;

Thornberry, 1973; Cohen and Kluegel, 1979a).

Being charged with more than one offense also increases the like-

lihood of a petition recommendation substantially. For those juveniles

with more than one offense charged against them at arrest, the prob-

ability of receiving a petition recommendation, net of other factors,

is 17.2 percentage points higher than for youths who are charged with

only one offense. Previous research on intake decision-making has not

generally included a similar measure, and these results suggest that

number of charges may be an important consideration in intake decisions.

It is difficult to know exactly why number of charges affect the prob-

ability of a petition recommendation, but it may be that this factor

is used by intake workers as a measure of a juvenile's culpability.

That is, being charged with more than one offense at arrest may

indicate to intake workers that the juvenile is somehow more committed

to delinquency than a youth who is charged with only one offense.

Alternatively, the number of charges may be interpreted as additional

information about the seriousness of the offense in question.

Of all the sociodemographic variables, the age of the juvenile

has the strongest impact on intake recommendation. An increase from

one age category to the next raises the probability of a petition

recommendation by nearly nine percentage points (8.8%). This means

that a juvenile who is age 16 or older is 17.6 percentage points more

likely than a child under age 13 to be recommended for a petition to

court. Studies by Terry (1967) and Thomas and Sieverdes (1975) report

similar patterns. However, none of the studies that introduced

simultaneous controls for all independent variables included age as

a variable in their analyses. Therefore, this finding becomes an

important addition to the empirical literature, adding some support

for the contentions of theorists such as Lemert (1951), Turk (1969),

and Quinney (1970) about the relationship between age and justice

system processing.

Being black rather than white also produces a substantive change

in the likelihood of receiving a petition recommendation. Black youths

are 12.3 percentage points more likely than whites to be recommended

for a petition to court when other important independent variables are

controlled. Similar findings have been reported in some earlier work.

For example, Thomas and Sieverdes (1975) have shown significant race

effects on intake dispositions, controlling for one variable at a time.

Thornberry (1973) controlled for the effects of prior legal record and

offense seriousness simultaneously and found that race remained a

significant predictor of intake decisions. Other multivariate analyses

(Terry, 1967; Arnold, 1971; Cohen and Kluegel, 1979a), however, did not

find race to be significant when a larger number of controls were intro-

duced for legal and sociodemographic variables. The present findings

are opposite those reported in these latter studies. The present

findings show that, net of other factors, race does exert a substan-

tive impact on the type of recommendation that intake workers make.

Agency case status is the only legal variable that does not ap-

preciably alter the probability of petition recommendation in the

multivariate analysis. This variable has less than a 10 percentage

point (.09%) impact on the probability of petition recommendation.

This is the level selected as guides for this study, but others may

wish to argue that this difference is still worth noting. Even when

prior legal record is controlled, HRS status has some power to alter

the probability of a petition recommendation. The effect of agency

case status on intake recommendation is thus left open for others to

debate, and its effect at later stages of the juvenile justice system

will be evaluated in subsequent analyses.

A juvenile's sex has the smallest effect of any independent vari-

able entered in the logit model. Net of other factors, being female

rather than male does not meaningfully alter the probability of peti-

tion recommendation. This pattern is consistent with some other re-

search that has found minimal sex effects in multivariate analyses,

and poses some interpretive problems for authors who expect processing

patterns to vary by sex (e.g., Chesney-Lind, 1977; Cohen and Kluegel,


All in all, the results of this analysis are intriguing. With

the exception of agency case status, legal factors seem predictive of

intake recommendations, and are far better predictors than are socio-

demographic characteristics. Net of these legal variables, certain

sociodemographic variables do substantively alter the probability that

intake workers will recommend a petition to court, but their impact is

rather modest. Older juveniles and black youths are more likely than

younger and white juveniles to receive harsh petition recommendations,

indicating differential treatment at the intake stage.

The preceding results were obtained by evaluating the logit coef-

ficients at the mean of the dependent variable. This is the point at

which the effects of the independent variables should be greatest. It

was pointed out in Chapter Three that the impact of the independent

variables would diminish as the probability neared either end of the

curve. In order to observe the impacts of the independent variables

at other points, the probabilities have also been evaluated at two

other theoretically interesting points to see how the effects differ

at the opposite ends of the curve. First, the effects were evaluated

at the point where the values of all the independent variables are

zero. These are "minor" offenders who are younger, female, and white.

In addition, they are charged with misdemeanors and come into the

system with one charge, no prior record of agency contact, and no

prior legal record. Such a child would have only a 19.4 per cent

probability of being recommended for a petition, consistent with

theoretical expectations. Second, the effects were evaluated at the

other extreme, the theoretically most "serious" offender. That is,


an older, male, black, with prior HRS contact, a prior legal record,

a felony offense, and more than one charge at arrest. Such a juvenile

would have a 96.2 per cent probability of receiving a petition recom-

mendation. The predicted proportional changes for these groups are

shown in Columns D and E, respectively, of Table 4.3. In the present

discussion primary attention will be given to the effects of age and

race since they are the variables that suggest differential treatment

on the basis of sociodemographic factors at the mean value of the

dependent variable.

It can be seen that the impact of age is indeed smaller at either

end than at the mean, and the same statement holds for the effects of

each independent variable. The effect of race falls below the .10

change level used to signify substantive impact at the mean. Age has

its greatest effect at the "minor" offender end, producing an 11.4

percentage point change between the youngest and oldest juveniles'

probabilities of petition recommendation. At the other extreme, older

juveniles are only 7.6 percentage points more likely to receive peti-

tion recommendations that their younger counterparts. Race exhibits

a 7.9 percentage point difference between blacks and whites at the

"minor" offender lower end of the curve, but only a 5.3 percentage

point difference between these groups at the "serious" offender end.

While blacks are more likely than whites to be recommended for a peti-

tion at either extreme point on the logistic curve, the effects of

race are not great. Overall, it appears that age and race have their

greatest impact on minor cases, rather than on serious cases. With

regards to each of the other independent variables, they too have their

stronger effect at the lower end of these two points. Thus, it may be

that at the "minor" end the relative unimportance of the offense leads

the intake workers to overlook the other factors in the case. At the

"serious" end, though, sociodemographic factors may have relatively

little impact, because the overall seriousness of the case may practic-

ally dictate the decision to be made. There are still differences be-

tween age and racial categories at either extreme of the curve, but

with the exception of the impact of age at the "minor" end, they seem

to be fairly small and do not allow any clear statement about the dif-

ferential treatment of these types of cases.

With respect to the central questions guiding this research, there

are some specific points to be discussed. Both legal and sociodemo-

graphic factors play important roles in determining intake decisions.

Offense seriousness, prior legal record, and number of charges are the

best predictors of intake recommendation. The effects of age and race

on intake decisions, however, are not far behind number of charges.

The fact that age and race are predictors of intake dispositions does

suggest the possibility of differential treatment of less powerful so-

cial groups. Older juveniles and blacks, as predicted by conflict and

societal reaction theorists, are found to be more severely sanctioned

than are young or white offenders at the intake stage. Even when the

effects of other variables are taken into account, these relationships

persist. If one assumes all relevant factors have been controlled

through these analyses, then the present results may be seen to be

consistent with the contentions of conflict and societal reaction

theorists concerning the application of the law. The findings, over-

all, are also consistent with much of the previous research on intake

dispositions. Differences between these findings and those in the

literature may be explained in several ways.

First, with the exception of Cohen and Kluegel's (1979a) log-

linear analysis, the analytic technique utilized here is more compre-

hensive than those used in most previous research. Second, these data

are taken from a more recent time period than any other published data.

Earlier research that shows differential treatment patterns by socio-

demographic categories may reflect historical patterns of treatment

that have since changed, which would suggest that differential treat-

ment on the basis should diminish. Third, the data are drawn from a

county in a Southern state, although it is primarily an urban county

with a heterogeneous population. To the extent that treatment prac-

tices vary from jurisdiction to jurisdiction, different predictions

of disposition can appear in analyses using different data sets.

There are, however, similarities between these findings and those re-

ported in other research that used data from other regions of the

country (e.g., Goldman, 1969; Thornberry, 1973). Further, similar

patterns have been reported in jurisdictions where the intake process

is handled by the police (Goldman, 1969; McEachern and Bauzer, 1967),

or by the court itself (Thornberry, 1973; Thomas and Sieverdes, 1975),

as explained in Chapter Two. The generalizability of the present

findings will be addressed in the discussion in Chapter Five.

Juveniles with prior legal records, as shown in Table 4.1, are

much more likely to be recommended for a petition to court than those

without a prior record. Seventy-seven per cent of those with prior

records receive petition recommendations, while only 55.45 per cent

of those with no prior record receive such a recommendation. This is

consistent with the results of previous empirical research that has

considered prior record as an independent variable (McEachern and

Bauzer, 1967; Terry, 1967; Arnold, 1971; Thornberry, 1973; Thomas and

Sieverdes, 1975).

In the case processing literature offense seriousness is consist-

ently associated with intake decisions. The present findings prove no

exception to this rule; offense seriousness is associated with the type

of recommendation that the intake workers make. As shown in Table 4.1,

a slim majority of those charged with misdemeanor offenses (53.49%)

are given no petition recommendations, but nearly three-quarters

(74.85%) of those charged with felony offenses are recommended for a

petition to court. The chi-square test indicates that this association

is statistically significant.

Finally, the number of charges at the time of arrest is also as-

sociated with the type of recommendation made by the intake worker.

While a majority (57.19%) of those charged with only one offense re-

ceive petition recommendations, a substantially greater proportion of

those charged with more than one offense (78.13%) are given petition

recommendations. This relationship is also statistically significant.

State Attorney's Action

So far the substantive impacts of both legal and sociodemographic

variables on intake recommendation have been examined. Legal factors

appear to exert the greatest overall influence on the type of recom-

mendation given, but two of the three sociodemographic factors also

show substantive relationships. These patterns suggest differential

treatment of blacks and older youths who appear to be handicapped in

the first stages of their experience with the juvenile justice system.

It must be remembered, however, that the intake recommendation is

simply that--a recommendation to the State Attorney. The State At-

torney does not have to follow the recommendation. The zero-order

correlation between intake recommendation and State Attorney's action

in these data is .35. This suggests a tendency for the State Attorney

to accept the intake worker's recommendation, but the relationship is

not strong. The final decision on whether or not to petition a child

to court, then, rests with the State Attorney. This raises two ques-

tions: 1) what factors are related to the action that the State At-

torney takes; and 2) whether intake recommendations themselves have

any effect on State Attorney's actions, net of the other independent


In order to examine these questions, two logistic regression

models have been constructed.3 The first model regresses State Attor-

ney's action on the seven independent variables used in the previous

analysis of intake recommendation. The second model incorporates the

intake recommendation as an additional independent variable. Both

models are evaluated at the mean value of State Attorney's action.

The results of these analyses are presented in Table 4.4.

In Model 1, only four of the coefficients display predicted pro-

portional change values greater than 10 per cent. Offense seriousness

has the strongest effect on State Attorney's action. Youths charged

with felonies are 24.1 percentage points more likely to have a peti-

tion filed against them than are those charged with misdemeanors.

Juveniles with prior legal records, as opposed to those with no prior

records, have an 18.3 percentage point greater probability of being

filed on. An increase from one age category to the next produces a

9.1 percentage point difference in petition likelihood. Those youths,

age 16 or older, are thus 18.2 percentage points more likely to be

petitioned than youths under the age of 13. Being charged with more

than one offense at arrest raises the chance of petition by 13.6 per-

centage points over the effect of being charged with only one offense.

The impact of the juvenile's race, sex, and agency case status is

negligible. It is interesting to note that current HRS clients are

no more likely to be petitioned than are new clients, suggesting that,

unlike its impact on intake workers, this factor plays no role in the

State Attorney's decision.

Model 2 incorporates intake recommendation into the logit equa-

tion as an independent variable. Intake recommendation shows the

strongest predictive impact of any independent variable on State At-

torney's action. Net of legal and sociodemographic variables, being



-0 -

)O 0











D Cd" o C) o C --- MQ
LO o0 -- -- -- LO

I I-
co Lo CMJ rl r-. o o0

3- 0 C0 O iV CY-

lz 0 C )

3 S- QU) .

4- o e. .-
t)-o u -V
cu x u 4-) a
Cm () (a c(I o i

(A ( (n) V) o) 0)

- I- -j a)
*r *r S-
C C U C a)
0) a) a) C 0 a -Q
> > > 4- E
3 3 s- 4- 3
S3 < 0 Z

I r-

i Cm LA
I Kr L
I C *tj-




4- 4-)
0. .

-i 1-

c-, i-- o c' od" OO cM 0"
- C CV K 00 r-.
0 0 0 0 0 0

r-- p'J i- cOJ rJ cMJ f. n- r
i- CJ C0 j C C0 C M j

I I I -

| 0 CJ i- o N
e e a s e *C~
brO C~~~bQ IC Q0On




4-- LO)
Sr- t

S0. r --

U C 0
r_ 0

4- 4
,a n

5 '.r

- 0
n C

*-- -.- I
a 0 4-
*O O 0

i i x
4- CA

1 0 4
^ o 0

4- 4- *


recommended for petition by the intake worker increases the likelihood

of actually being petitioned by the State Attorney by 27 percentage

points. There are still other variables that meaningfully alter the

probability of petition even when intake recommendation is controlled.

Being charged with a felony increases the probability of petition by

18.9 percentage points over being charged with a misdemeanor. Having

a prior record raises the likelihood of a petition by 14.9 percentage

points over having no prior record. Being charged with multiple of-

fenses versus one offense increases the likelihood of petition by

12.4 percentage points. Finally, being over 16 versus under 13 years

of age produces a 14.6 percentage point greater probability of peti-

tion. The juvenile's sex, race, and agency case status show little

effect on the probability of petition by the State Attorney.

Judicial Decision

Intake recommendations do appear to have a substantive impact on

whether or not the State Attorney decides to file a petition to court.

Once a petition to court has been filed, does this influence continue?

Further, do the other independent variables in the previous models in-

fluence decisions made at the judicial level of the juvenile justice

process? In order to address these questions, judicial decision was

regressed on the independent variables. There were three categories

of judicial decision that could be utilized (as described in Chapter

Three). Since the differences in severity between each category are,

more or less, equal, ordinary least squares (OLS) regression techniques

were employed in this analysis. Model 1 of Table 4.5 contains the

results of regressing judicial decision on only the seven independent

variables used in the analysis of intake recommendation. Model 2 in-

corporates intake recommendation as the eighth independent variable.

An examination of Model 1 shows that four variables that have

been defined as substantive predictors of intake recommendations and

State Attorney's actions also influence judicial decisions. Offense

seriousness has the strongest effect (B = .353) on judicial decision,

with prior legal record a fairly distant second (B = .170). The impact

of a juvenile's race appears to be substantial (B = .123), though not

statistically significant. The remaining independent variables exert

minimal predictive power on judicial decision.

Model 2 of Table 4.5 reveals that the impact of these variables

is diminished somewhat with the addition of intake recommendation into

the equation. The order of importance remains more or less the same,

with intake recommendation occupying the third position. As in Model

1, offense seriousness has the strongest effect on judicial decision

(B = .329), with prior legal record again a distant second (B = .155).

The introduction of intake recommendation into the model shows it to

exert a weak influence (B = .120). The impact of race as a predictor

drops somewhat from Model 1 (B = .116). The effects of the remaining

variables are again small, but they are in the same directions as

noted in Model 1. It appears that legal case characteristics play the

most important roles in judicial decision-making, though the effect of

race and intake recommendation appear to be worth noting.

0 t tO Ol C0 C

I 0 C .-




V )

o .-



oU -o


S- C'ID 0 C 0 to
C 0 0 V 0


0 o 0 Io ,
o w 0 *


0m1 o r- -z CM
0 0Ln m 01 o r

-0 Ln

S- 0 C

U ) ) V 0U

OC 0 Ln (A t 0V
U 0 0 0

>) >) U () (U
*7- *rS *f
C C C U S- C

> > > ) r- 4-
3 3 3 01 I1 4-
-> -) CL 0




Q) 4L
) L-
rm U)
-^ s
C r
0-4 0-

n m 00o0C C\j r% oC
o 0 O -

OD CO U-) tUb M> q* (M -4 O 00
qc t o c1 Kr c Ij d C\j LO r
O C 0 %\j LO C\j 0 C'j



(a C\-



4- U
4- -
U 0
o s


Intake recommendations may be viewed as playing a substantive

role in the State Attorney's decision on whether or not to petition

a youth to court. In addition, they play a minor role in the final

judicial decision. Intake recommendations thus exert an impact on

the treatment that juveniles receive throughout the juvenile justice

system. These findings are similar to those reported by Hagan (1975)

in the criminal justice system. In the final chapter, the influence

of the nonlegal sociodemographic variables on intake recommendation,

and the remainder of the juvenile justice system, will be discussed

as it relates to the issues of differential treatment and discrimina-

tion in that system.


Agency case status and prior legal record show the strongest
correlation (r = .66) between any two variables in the matrix, and
the only relationship that might warn of multicollinearity. While
these two variables may be measuring aspects of the same phenomenon,
agency case status is a measure of contact with the agency handling
intake within a relatively short time period, and an indication that
a previous case involving the juvenile has not been closed. This con-
tact may be for either a delinquency charge, or an abuse/neglect or
dependency incident. Prior legal record, on the other hand, is an
explicit measure of prior involvement with HRS intake on a delin-
quency referral.

2An analysis of the missing cases was performed to determine
whether there were substantial differences between them and the re-
maining cases on values of the independent variables. No discern-
able patterns were noted. Therefore, it was decided that they should
be treated as missing data.

3The State Attorney's action was recorded only on the second
Intake Data Card in a manner that allowed the author to analyze it.
For this reason, there are only 456 total cases available for anal-
ysis. It can be seen from Table 4.4 that there was little missing
data. State Attorney's action was recorded in the same fashion as
intake recommendation, and since it was coded into a dichotomous
variable, logistic regression was used.




This study began with personal curiosity about how intake workers

decided what children should or should not be referred to a juvenile

diversion program with which the author was associated. It grew to

encompass the general issue of juvenile intake decision-making after

encountering the literature on juvenile justice processing. There was

also the fact that more data on general intake decisions were at the

author's disposal than data on any other aspect of juvenile justice

decision-making. In exploring the area of decision-making in the ju-

venile (and criminal) justice system, the issue of differential pro-

cessing became more interesting. Examining the impact of sociodemo-

graphic and legal factors on intake counselors' decisions, particularly

after talking with these people, seemed to be a natural area of study.

The issue presented not only a practical problem to examine, but a vast

theoretical literature that was both complex and interesting.

The reader will remember, with respect to the theoretical issues,

that the "moral functionalists" (Hagan, 1980) argue the law is applied

primarily on the basis of legal considerations, and that nonlegal,

particularly sociodemographic, factors play little or no role in jus-

tice system decision-making. The "moral Marxists" (Hagan, 1980) and

the societal reaction theorists, on the other hand, assert that after

legal considerations are controlled, sociodemographic characteristics

indicative of an individual's social power play a substantive role in

justice system decision-making. Previous empirical research on this

subject, in both the criminal and juvenile justice system, however,

has proven inconclusive.

In order to address the issues outlined above and in the earlier

chapters, intake records for 1,209 delinquency referral cases from one

Florida county were analyzed using logistic regression techniques.

These records covered an 18 month period and contained legal, socio-

demographic, and outcome information relating to each case. While

the main focus of this study was the intake decision point, the impact

of the legal and sociodemographic variables on subsequent decision

points was also examined. In addition to the effects of these inde-

pendent variables alone on each decision point, the additional impact

of the intake recommendation on the later decision points was also


A Methodological Note

The use of quantitative techniques to study juvenile intake deci-

sion-making (and for that matter, any justice system decision-making),

has come under repeated attack from some analysts (Kitsuse and Cicourel,

1963; Cicourel, 1968; Needleman, 1981). These critiques center around

two issues: First, the assumptions that researchers make about the jus-

tice process when utilizing official statistics; and, second, the mean-

ing of the statistics themselves and the interpretations given to them

by researchers.

In a recent study of intake decision-making, Needleman (1981: 248)

summarizes these issues by focusing on some basic though implicit,

assumptions about the intake process held by most researchers, assump-

tions which she asserts may not be valid:

1. Being sent to court is necessarily a harsher sanc-
tion than being diverted from court. (Thus sending a
case to court indicates a more negative evaluation on
the part of the intake worker than does a diversion
from court.)
2. Screening decisions necessarily involve judge-
ments about the individual juvenile offender. (Thus
personal characteristics of those sent to court may
indicate which juveniles tend to evoke negative eval-
3. Despite their artifactual nature, official records
do capture with reasonable accuracy the basic facts of
how cases are being handled. (Thus official records
are a useful data source for studying the processing
of cases.)

Needleman argues that her own experience in observing intake pro-

cedures suggests these assumptions to be false. Her criticisms are

germane to the present study, and will be examined separately as they

relate to this study.

Needleman (1981: 250-253) writes that the intake workers she ob-

served did not view court referral as a necessarily harsher penalty

than those that they could impose themselves. The intake workers she

studied could not only make dispositions about further processing, but

could also intervene directly into the juvenile's life through a pro-

vision for informal probation. This is an important distinction be-

tween the two juvenile intake units being studied in the respective

studies. In Florida, intake workers have little power to intervene

in the juvenile's life. They can recommend some relatively mundane

sanctions (e.g., a letter of apology to the victim), usually in antici-

pation of the State Attorney's actions, but these are strictly on a

voluntary basis with the cooperation of the client. In addition,

Florida intake workers only make recommendations concerning court

petition and dispositions; they do not have the authority to determine

final case dispositions. Finally, intake workers interviewed by the

author, though at a later date than the study period, did view court

as a generally more serious sanction than any form of noncourt action.

Thus, some important differences appear to exist between the practices

and attitudes of the intake workers in the two different locales stud-

ied by Needleman (1981) and by the present author.

Needleman's (1981: 253-255) second point is not so easily dis-

missed. This study has assumed that characteristics of the individual

do have an impact on intake recommendations. This does not mean that

the social situation of the juvenile is not presumed to play an import-

ant role in the decision-making, nor does it imply that personal char-

acteristics are not correlated with the social situation of the juven-

ile in the minds of the workers. Rather, there simply are no measures

of relevant social situations contained in these data, situations that

researchers have found to play important roles in decision-making (e.g.,

Cicourel, 1968; Horowitz and Wasserman, 1980). In defense of the pre-

sent study, however, it is argued that in at least some cases, it is

the juvenile that is the target of the intake worker's decisions,

rather than the situation in which the juvenile lives. This is not

to say that even in the case where the youth is the center of atten-

tion that the social situation is not a potential consideration.

Rather, social factors are important, but there are many times when

the intake workers assess responsibility on the part of the juvenile,

independent of social circumstances.

Finally, Needleman (1981: 255-260) charges that official records

data are often misleading, if not false in their portrayal of the

intake procedure. This criticism is one of the most difficult to

deal with if one argues that official data are accurate reflections

of the processes at work in the decision-making situation. In defense

of the present study, and the framework developed here, this is not

necessarily the case. To the extent that the variables employed here

are not subject to a great deal of interpretive work on the part of

intake workers, it is assumed that these data do present an accurate

representation of some of the legal and sociodemographic facts of a

case that might be taken into consideration by the intake workers.

Needleman (1981: 256) writes that a form of "charge bargaining" went

on in the intake units she studied. The original charge from the

police and the charge that appeared on the intake record were often

different due to manipulations and negotiations on the part of the

intake workers. Intake workers in Florida do not set charges. The

charge that appears on the record is the police charge at arrest,

which has also been forwarded to the State Attorney's office. It is

the State Attorney who has the authority to alter charges, not the

intake worker. To the extent that the variables used here are not

subject to much manipulation on the part of intake workers, they ap-

pear to accurately reflect certain realities of the cases. There is