dee

Delinquency as the Failure of Personal and Social Controls

Author(s): Albert J. Reiss, Jr.

Source: American Sociological Review , Apr., 1951, Vol. 16, No. 2 (Apr., 1951), pp. 196- 207

Published by: American Sociological Association

Stable URL: https://www.jstor.org/stable/2087693

REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/2087693?seq=1&cid=pdf- reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to American Sociological Review

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

196 AMERICAN SOCIOLOGICAL REVIEW

white collar group is over-represented at Indiana University.

3. The farming, semi-skilled and un- skilled groups are strongly attracted by the teaching profession.

4. Students from the farming and blue collar groups stand a better chance of having their college education postponed or inter- rupted by a work experience than students coming from the white collar group.

5. Students from the white collar group stand a better chance of becoming members of social fraternities at Indiana University than students coming from the farming and blue collar groups.

6. The G. I. Bill of Rights has increased the proportion of students from the blue collar group at Indiana University from 90 to 113 per cent.

CONCLUSIONS

1. The white collar group was found to be over-represented not only among the non- veterans but also among the veterans at In- diana University. If the assumption is granted that the post-war veterans who

entered the military indirectly are the group most likely to contain students who would otherwise be unable to attend the university, the corollary follows that many white collar families, out of proportion to their numbers in the state, are unable financially to send their sons to the university. This is not very surprising when one considers that the an- nual median family income in the United States in 1947 was only $2,800.

2. The farming, semi-skilled and un- skilled groups did not or were unable to take advantage of the G. I. Bill of Rights in proportion to their numbers in the state of Indiana.

3. On the basis of the above parochial study it may tentatively be hypothesized that, in general, the absence of talented students from the white collar and skilled groups in institutions of higher learning is due, on the whole, to economic factors rather than to cultural factors, but that, in general, the absence of talented students from'farm- ing, semi-skilled and unskilled groups in institutions of higher learning is due, on the whole, to cultural factors rather than to purely economic factors.

DELINQUENCY AS THE FAILURE OF PERSONAL AND SOCIAL CONTROLS*

ALBERT J. REISS, JR.

University of Chicago

D ELINQUENCY may de defined as the behavior consequent to the failure of personal and social controls to pro-

duce behavior in conformity with the norms of the social system to which legal penalties are attached. Personal control may be de- fined as the ability of the individual to refrain from meeting needs in ways which conflict with the norms and rules of the community. Social control may be defined as the ability of social groups or institutions to make norms or rules effective.

*Paper read at the annual meeting of the American Sociological Society held in Denver, Sep- tember 7-9, 1950.

Delinquency results when there is a rela- tive absence of internalized norms and rules governing behavior in conformity with the norms of the social system to which legal penalties are attached, a breakdown in pre- viously established controls, and/or a rela- tive absence of or conflict in social rules or techniques for enforcing such behavior in the social groups or institutions of which the person is a member. Hence delinquency may be seen as a functional consequence of the type of relationship established among the personal and social controls.' While for

1 This formulation is not in contradiction with formulations which view certain types of delin-

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

FAILURE OF PERSONAL AND SOCIAL CONTROLS 197

a particular person the relative weakness of personal or social controls may account in large part for the delinquent behavior, for most cases the relative weakness of personal and social controls should account for the delinquent behavior. Similarly, when we con- sider the problem of delinquent recidivism, i.e., continuation in delinquent behavior, we would expect that personal and social controls are less effective in governing the behavior of delinquent recidivists than non- recidivists.

The purpose of the present paper is to isolate a set of personal and social controls which are associated with delinquent re- cidivism and to evaluate them as prognostic of recidivism. The evaluation of the predic- tors is undertaken in terms of the theory of prediction.2

The Predictors and the Criterion. The data collected for this study consist of the materials in the official juvenile court records of 1,110 white male juvenile delinquent probationers between the ages of 11 and 17. These delinquents were placed on probation by the judges of the Cook County Juvenile Court between March 4, 1943 and October 31, 1944. From the data in the first 736 case folders, items were selected which are measures of personal controls, social control in primary groups, and social control by community and institutions. Factor items were then isolated as predictors in terms of the theory of prediction and a prediction device constructed. Finally, predictions were validated by predicting for a follow-up sam- ple of 374 cases.

quency as a consequence of social control in the delinquent gang. The delinquent peer group is here viewed as a functional consequence of the failure of personal and social controls in the social system to produce behavior in conformity with the norms of the social system to which legal penalties are at- tached. Acceptance of the delinquent gang as an agency of control results in rejection of the con- formity norms of the larger social system or the acceptance of and/or submission to conflicting sets of norms and rules.

2 This paper summarizes a part of "The Accu- racy, Efficiency and Validity of a Prediction Instru- ment" (unpublished Ph.D. dissertation, Department of Sociology, University of Chicago, 1949).

A major consideration in prediction studies is the character of the event predicted. The criterion predicted in this research is official action. It is a dichotomous classification of success or failure on probation of delin- quents from the Cook County Juvenile Court. In effect, most probationers (92%) classified as failures commit a delinquent offense while under probationary supervi- sion. Eighty-seven per cent of these offenses consist of adult felonies. Since over ninety per cent of the delinquents who violate pro- bationary supervision commit offenses, it seems logical to conclude the behavior of the delinquent determines official action. In terms of the theory of delinquent recidivism, then, our assumptions are twofold: (1) a delinquent's behavior on probation deter- mines action by official authority3 and (2) the observed association between predictors and criterion is valid in the sense that it is prognostic of delinquent recidivism.4

3 There is the possibility that all delinquent pro- bationers who commit an offense are not included in the failure population. It is doubtful whether such selection significantly alters the observed relation- ships in this research. If it is argued such selection significantly alters the relationship, it must be dem- onstrated that the nature of official action has materially altered the association of predictors and the criterion. (This does not negate the possibility that the nature of official action may attenuate the observed correlations). Previous studies demon- strate that when official selection operates to "select" a population, it may operate with respect to the factors associated with the behavior predicted. Hence official selection into a population does not necessarily alter the relationship between a set of factors and the behavior predicted. See United States Department of Justice, Attorney General's Survey of Release Procedures, Washington, D.C.: United States Government Printing Office, 1939, Vol. IV, Appendix A.

4 To the extent this assumption cannot be

granted, we are of course predicting official action

rather than a delinquent's behavior. The assumption

does not preclude the assignment of changes in the rate of violation over periods of time to official

causes, e.g., changes in legal norms, law enforcement policy or probation supervision. Furthermore, varia- tion in the nature of official action may affect indi- vidual probationers or classes of probationers differ- entially. Over short periods of time, however, a relatively consistent rate of violation among the classes is generally maintained

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

198 AMERICAN SOCIOLOGICAL REVIEW

CONTROL IN PRIMARY GROUPS

Primary groups are the basic institutions for the development of personal controls and the exercise of social control over the child. These groups exercise social control over the non-delinquent child by providing non-de- linquent social roles and by employing tech- niques which make non-delinquent norms

and rules effective. Concomitantly, the non- delinquent child develops contra delinquent personal controls in primary groups. Delin- quency and delinquent recidivism may be viewed as a consequence of the failure of primary groups to provide the child with

appropriate non-delinquent social roles and to exercise social control over the child so these roles are accepted or submitted to in accord with needs.

Meeting the Needs of the Child. An im- portant part of family control is the ability of the family to meet the needs of its mem- bers. An aspect of this is the ability of the family to provide for member needs through the purchase of material goods and services. In Table 1, we observe there is a low but statistically significant association between income or economic status of family and outcome on probation. The child of the economically dependent family, however, constitutes a significantly poorer probation risk than the child of a family with earned income.

The economically dependent family is one where the impaired financial ability may often result in objective deprivations to members. Economic deprivation may lead to subjective feelings of insecurity about the possibility of future needs being met and therefore weaken personal controls. Further, the disapproval of a society which fails to provide for one's needs is of less concern when a person's position is one of economic dependence upon the society than one of relative independence. Finally, dependent economic status often results in diminished parental authority over family members. Loss of parental authority (particularly that of the father) is greatest over the adolescent members of the family.5 Since economic de-

5 See Mirra Kormarovsky, The Unemployed Man

pendency appears to have marked disintegra- tive effects upon family unity and control and the child's acceptance of social control, adolescents from these families could be expected to meet their needs more often through delinquent activity than members of families who are not similarly deprived.

Dissemination of Norms and Rules. The child develops appropriate personal controls and the family exercises contra delinquent social control over the child's behavior when the family milieu is structured so that the child identifies with family members who represent roles of conformity with non-de- linquent norms and accepts the norms and rules embodied in these roles. Such families are a social-psychological unity oriented toward the establishment and maintenance of non-delinquent behavior for its members.

Studies of recidivism in delinquent popu- lations show that delinquents from families where there is a formal break in the struc- tural unity tend to be recidivists more often than delinquents where parents are living together. Our observations on marital status

of parents in Table 1 are in accord with these conclusions. Absence of a formal break in the structural unity of the family does not necessarily mean, however, that the social-psychological unity of the family is intact. Such considerations necessitate an examination of the nature of the formal break vis-a-vis the social psychological unity of the family. Table 1 shows that delinquents from homes where the psychological unity is impaired (as measured by the character of the marital relationships between parents at time of delinquency) are significantly more often recidivists than delinquents from homes where the psychological unity is rela- tively intact.'; A comparison of the data with

and His Family, New York: The Dryden Press, 1940. Similarly Cavan and Ranck observe that de- pendent economic status decreases the ability of the family to act together as a unit. Ruth S. Cavan and Katherine Ranck, The Family and the Depression, Chicago: University of Chicago Press, 1938.

6 The index of the social-psychological unity of the family employed here is the character of the marital relations between husband and wife. Fami- lies may be considered psychologically impaired as a unity when there is an open breach or gross in-

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

FAILURE OF PERSONAL AND SOCIAL CONTROLS 199

those on formal break in the structural unity of the family suggest that delinquents from homes where the psychological unity is im- paired but parents remain together are more often failures on probation than delinquents from homes where an open breach occurs.7

A major factor in the development of per- sonal controls is the moral ideal which par- ents represent and the techniques they use to control the child's behavior. If the child is to develop adequate non-delinquent controls, parents must represent social roles congruent with the non-delinquent norms and rules of society and control must be exercised so the child will accept these social definitions. The child will generally not accept non- delinquent norms and rules (even though represented by parental models) when par- ents over-control or under-control the child's behavior, e.g., when parents punish the child severely, neglect or abandon the child or are extremely lax or indulgent. Data on par- ental moral ideals and techniques of control over the child's behavior in Table 1 show that delinquents from families with unfavor- able moral ideals and/or techniques of con- trol are more often recidivists than delin- quents from families with favorable moral ideals and/or techniques of control.8 We also observe that children from families

compatibility between husband and wife. They are psychologically intact when there are relatively strong affective ties between spouses. In the present study an open breach is defined as permanent sepa- ration, desertion or divorce, intermittent separations or desertions, or appearance in a court of domestic relations. Gross incompatibility is defined as fre- quent and persistent disagreement and conflict.

7 One reason for this may be the fact that in homes where the formal structure remains intact but psychological unity is impaired, the child often be- comes the center of the conflict. Examination of the case records shows parents of these delinquents often disagree over discipline of the child, compete for the affection of the child, etc.

8 Parents were classified as having favorable moral ideals and/or techniques of control when they represented social roles congruent with non-delin- quent norms and rules and did not over-control or under-control the child's behavior. They were classi- fied as having unfavorable moral ideals and/or techniques of control when they did not represent social roles congruent with non-delinquent norms and rules and over- or under-controlled the child's behavior.

where a parent is absent but favorable moral ideals and/or techniques of control prevail are less often failures on probation than children from families where unfavorable moral ideals and/or techniques of control prevail.

It is generally agreed that children develop and maintain personal controls less readily when raised in a milieu other than the family of procreation. Children who spend part of their childhood in a children's institution are observed to be less able to control their own behavior and accept the control of other social institutions than the family child of the same age. While foster homes are usually observed to be inadequate for the develop- ment of strong personal controls (particu- larly as such placement is often conducted) the child is nonetheless a member of a family group with primary relationships and con- trols. In contrast, the institutional child is generally in an environment of deperson- alized norms and rules where the primary relationships and controls of the family are absent. These children not only feel manip- ulated by the impersonal milieu but usually fail to accept the norms and rules of the institution as personal controls. This does not necessarily mean open conflict with in- stitutional authorities. It may mean submis- sion to institutional authority.9 Data on institutional experience of the child in Table 1 show that delinquents who have experi- enced institutional placement are signifi- cantly more often recidivists than children who have had no such experience. While the number of children with foster home place- ment is too small to draw statistically reli- able conclusions, children from foster homes are apparently less often recidivists than children who have experienced institutional placement.

9 The psychological effects of most children's in- stitutions on the social and emotional development of the child has recently been summarized by Bet- telheim and Sylvester. Under prevailing standards of institutional treatment, the institution often has a psychologically crippling effect upon the achieve- ment of adequate controls in the child. See Bruno Bettelheim and Emmy Sylvester, "A Therapeutic Milieu", The American Journal of Ortho-psychiatry, 2 (1948), 191-206.

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

200 AMERICAN SOCIOLOGICAL REVIEW

TABLE 1. SMALL PRIMARY GROUP CONTROLS AS CORRELATES OF SUCCESS OR FAILURE OF JUVENILE DELINQUENTS ON PROBATION

Primary Group Controls Per Cent Total No. Critical As Prediction Factors1 Failure2 of cases Ratio3 T4

Family Economic Status** Comfortable .......................... 29.0 399 Marginal ............................ 33.8 257 ... Dependent ......................... 53.8 36 3.4 .12

Monthly Income of Parents* Over $270 ............................ 27.5 138 ... $180 to $270 ........................... 30.3 221 ... Less than $180 ......................... 34.2 243 Public Assistance ...................... 53.8 36 3.4 .09

Marital Status of Natural Parents* Parents Together ...................... 30.4 489 4.2 Separation, desertion, divorce ........... 37.3 142 ... One or both parents dead ............... 39.1 105

Marital Relationships of Natural Parents***

Strong affective ties ...................... 26.3 412 3.5 One or both parents dead ......... ........ 39.1 105 ... Open breach or gross incompatibility ....... 42.0 219 4.3 .13

Moral Ideals and/or Techniques of Control by Parents During Childhood of Delin- quent***5

Favorable moral ideals and/or techniques of control (parents together) .............. 26.3 410 4.2

Favorable moral ideals and/or techniques of control (1 parent absent) ............... 38.5 192 ...

Unfavorable moral ideals and/or techniques of control (parents together or one parent absent) ............................ 46.1 115 3.3 .13

Institutional or Foster Home Experience*** No previous experience ................... .30.4 670 4.2 Foster home experience ................... 41.2 17 Institutional experience .................. 64.4 45 4.7 .15

'Using a chi-squared test of significance these factors showed significant association with outcome on probation as follows: *.01<P(X2)<.10; **.001<P(X2)<.01; *** P(X2)<.001.

2The average violation rate of probationers is 32.7%. 8Predictors were selected by a technique of item analysis. A significant difference (C.R.>3) between

the mean violation rate of the sample and the mean violation rate of the sub-category of a prediction factor was used to select the items.

'Tschuprow's "T" was used to measure the degree of association between the prediction factor and outcome on probation.

'Nineteen cases were excluded from the analysis as they did not live in a parental environment during the greater part of their childhood.

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

FAILURE OF PERSONAL AND SOCIAL CONTROLS 201

COMMUNITY AND INSTITUTIONAL CONTROLS

A major source of a person's control lies in the social controls of the community and its institutions. From the standpoint of the group, control by such institutions lies in the nature and strength of the norms of the institutions and the effectiveness of the in- stitutional rules in obtaining behavior in conformity with the norms. From the per- spective of the person, institutional control lies in the acceptance of or submission to the authority of the institution and the reen- forcement of existing personal controls by institutional controls.

Sociologists have observed that delin- quency is more prevalent in certain neigh- borhoods and residential areas of the city. These areas are formally characterized as having a high degree of congestion or physi- cal overcrowding, limited recreational facil- ities and low-rental dwelling units in a poor state of physical repair interspersed among commercial and industrial facilities. Com- munity institutions in these areas are often at variance with the norms of individuals and groups in the community or limited in their capacity to meet individual and group needs. The norms of residents are observed to be relatively at variance with the norms and rules of the social system and the consensus order is relatively weak and ineffective in controlling behavior in accord with non-delinquent norms and rules.10 Persons in these communities less often accept or submit to the control of institutions embodying non-delinquent norms and rules. Data in table 2 show that suc- cess on probation is positively associated with residence in areas where contra de- linquent institutional controls are pre- sumably strong. In suburban areas and areas of low delinquency, in residential neighborhoods and in good residential com- munities more delinquents succeed on pro- bation than in areas of high delinquency,

10 For a formal statement of this position see Henry D. McKay, "The Neighborhood and Child Conduct," The Annals, 261 (January 1949), 32-41.

in commercial areas and in poor residential communities."

Community institutions are probably more effective in controlling the behavior of family members when influence is exercised over a relatively long period of time. Further- more, families with a relatively stable social organization are more likely to establish permanent residence in a community. Such families are usually more integrated into the community and accept the control of conventional institutions. In table 2 we observe that children from families who own homes are significantly less often fail- ures on probation than children from families who rent their homes.12 Similarly, we ob- serve that children from families who change their place of residence infrequently are less often failures on probation than chil- dren from families who reside in the neigh- borhood over a relatively short period of time.

A major set of institutions which exer- cise control over the person are those pro- viding formal education. Within the school, the child is provided with a set of learning experiences which are presided over by adult authority. Such experiences generally take place within a framework of controlling behavior in conformity with conventional norms. A measure of the degree of accept- ance or submission of the person to the control of the school is the regularity of attendance at school since frequent and con- tinued absence without permission of school authorities removes the child from the contra delinquent controls of the school and pro-

11 The low association of residential characteris- tics with outcome on probation may perhaps be at- tributed to the relative crudeness of these items as measures of the strength of community and institu- tional controls. Such attributes, however, would not be expected to yield a high order of differentiation as they do not take into account the relationship between the person and the social controls.

12 Home ownership does not vary directly with class variables in Chicago. The proportion of home ownership is greater for low and high income groups than for middle income groups. For evidence on the stabilizing effect of home ownership on the modern urban family see Lillian Cohen, "Family Character- istics of Home Owners," American Journal of Soci- ology, 55 (1940), 565-71.

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

202 AMERICAN SOCIOLOGICAL REVIEW

TABLE 2. COMMUNITY AND INSTITUTIONAL CONTROLS AS CORRELATES OF SUCCESS OR FAILURE OF JUVENILE DELINQUENTS ON PROBATION

Community and Institutional Per Cent Total No. Critical Controls as Prediction Factors1 Failure2 of cases Ratio8 To

Residence in Delinquency Areas*** Suburban .29.9 87 Low delinquency .30.7 401 Average delinquency ....... ...... 34.5 177

High delinquency .44.8 67 ... .10

Residence in Neighborhood Areas**

Residential .29.8 450 ... Semi-residential .33.6 140

Industrial ..... 29.2 24 ... Commercial .44.6 83 2.5

Type of Residential Community*

Poor .37.0 284 ... Fair. 30.3 178

Good .27.8 234 ...

Home Ownership*

Owner .26.0 192 ... Renter .37.3 500 ... .08

Less than three years .38.9 267 3.2 Three to ten years .28.7 293 ... Over ten years .24.6 126 ... .12

Truancy Record***8 Seldom or never truant.................. 25.2 230 4.6 Frequently truant ...................... 40.9 220 2.0 Usually truant .57.3 68 3.9 .19

Deportment Record***' Excellent or good........................ 36.2 218 ... Fair .. ..... . 31.2 164

Poor .. ... 46.3 82 ... Very poor .56.5 23 2.6 .11

1 See footnote 1, Table 1. 2 See footnote 2, Table L.' See footnote 3, Table 1. See footnote 4, Table 1. ' Residential mobility is expressed as number of years at present address. ' All cases of delinquents who were permitted to withdraw from school (183 cases) are excluded from this analysis.

hibits exposure to a situation in which rational controls are developed and strength- ened.'3 Furthermore, while truancy may be considered a mode of adjustment to a situa- tion which is not accepted by the person,

13 The reasons for such absence are, however, varied and complex. For an extended discussion of the factors related to truant behavior, see National Society for the Study of Education, "Juvenile De- linquency and the Schools," The Forty-Second Yearbook, op. cit., Chapters I to V.

from the perspective of control, truancy is generally without consent of authority. In addition, the truant usually does not accept or submit to the control of other socially approved (non-delinquent) institutions dur- ing his period of truancy. He generally plays truant alone or with other truants. In table 2 we observe that truants are significantly more often recidivists than are non-truants.14

14 It is recognized that truancy is a violation of

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

FAILURE OF PERSONAL AND SOCIAL CONTROLS 203

Within the truancy group, however, persons who are usually truant are recidivists sig- nificantly more often than persons who are frequently truant.15

Another mode of perceiving the delin- quent's acceptance of the school as a control institution is the extent to which the child conforms to the discipline requirements of the classroom situation. Presumably the child who conforms to the conduct requirements of the classroom may be thought of as accepting or at least submitting to the con- trol of the school. The child who fails to conform to the situation, i.e. one who con- stitutes a behavior problem to the classroom teacher and school authorities, may be thought of as rejecting such control. In Table 2 we observe that a significantly higher proportion of delinquents classified as behavior problems by school authorities (the very poor) and delinquents demon- strating poor deportment are failures on probation than delinquents viewed by school

officials as having fair, good or excellent deportment.

PERSONAL CONTROLS

The nature and strength of personal con- trols are an index of the person's definition of how he will act in certain situations. It is generally assumed that the personality of individuals who conform to norms to which legal penalties are attached is characterized

by (a) mature ego ideals or non-delinquent social roles, i.e. internalized controls of so- cial groups governing behavior in conformity with non-delinquent group expectations, and (b) appropriate and flexible rational con-

the legal norms of the social system requiring regular attendance in school and is therefore technically classifiable as delinquency. This should not preclude its use however as a measure of rejection of social control for the prediction of continuation in delin- quent behavior. The tautology appears justified here as absence from school without permission of school authorities represents rejection of authority and controls contra all forms of delinquency.

15 "Frequently truant" is defined as absent from school without permission for two-fifths or more of the total instruction period; "usually truant" is defined as absent without permission, but in attend- ance for at least three-fifths of the total instruction period.

trols over behavior which permits conscious guidance of action in accord with non-delin- quent group expectations. Satisfaction of needs is not postponed until they may be fulfilled by conforming to such expectations.

Data on personal controls were classified into three major categories: (a) relatively strong ego and/or super-ego controls, (b) relatively weak ego controls (immature per- son or low self-esteem), and (c) relatively weak super-ego controls ("socially deterio- rated" or "delinquency oriented" person- ality).16 Sixty-six per cent of the delinquents were classified as having relatively strong personal controls. On the basis of psychiatric diagnosis, the remaining thirty-four per cent were classified as having relatively weak ego and/or super-ego controls. Table 3 shows that persons classified as having rela- tively weak ego and/or super-ego controls are found in a population of recidivists sig- nificantly more often than persons classified as having relatively strong personal controls (T=.23).

The recommendations for treatment of the delinquent made by court staff workers and IJR psychiatrists provide a second index of the nature and strength of personal controls. Home and community placement is recommended for all delinquents where, in the judgment of the case worker or psy- chiatrist, the delinquent has relatively strong personal controls. When personal controls are judged as relatively weak but intensive supervision in an environment with strong social controls may lead to rehabilitation, the psychiatrist recommends such supervi- sion with provision for a short period of institutionalization and therapy or foster placement either (a) prior to community placement or (b) contingent on case prog- ress. Finally, for those delinquents where the diagnosis is "marked social deteriora- tion" or "very immature personality devel- opment," the psychiatrist recommends either

16 Data on the nature and strength of the delin- quent's personal controls were obtained from the judgments of the social work staff at the juvenile court and the reports of the psychiatrists at the Institute for Juvenile Research (IJR), an agency providing psychiatric diagnosis and treatment.

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

204 AMERICAN SOCIOLOGICAL REVIEW

intensive psychotherapy with the individual remaining in the community or treatment within a closed institution. Examination of the data on recommendations for treatment shows that delinquents with relatively weak personal controls, for whom placement with intensive supervision in an environment with strong personal controls is recommended, are more often failures on probation than

TABLE 3. PERSONAL CONTROLS AS CORRELATES OF SUCCESS OR FAILURE OF JUVENILE

DELINQUENTS ON PROBATION

Personal Controls as Per Cent Total No. Critical Prediction Factors' Failure2 of Cases Ratio' To

Adequacy of Personal Controls Relatively strong ego and su- per-ego controls (IJR+no IJR) ....................... 22.2 492 6.5 Relatively weak ego controls (IJR) ...................... 51.6 159 5.7 Relatively weak super-ego con- trols (IJR) ................. 52.6 95 4.4 .23

Recommendations for Treatment Home and community trial

(IJR+no IJR) ............. 22.4 497 6.4 Community Trial with provi- sional placement or institu- tionalization (IJR) .......... 47.2 125 3.8 Closed Institution or psycho- therapy (IJR) .............. 61.4 114 7.1 .24

'Using a chi-squared test of significance these factors showed significant association with outcome on probation, P (X2) <.001.

2 The average violation rate of probationers is 32.7%. 'See footnote 3, Table 1. 4 See footnote 4, Table 1.

delinquents for whom home and commu- nity placement is recommended. Significantly poorer risks, however, were delinquents characterized as having relatively weak ego or super-ego controls and in need of inten- sive psychiatric treatment in a community or institutional setting. Sixty-one per cent of these cases were found in the popula- tion of failures on probation.

Our hypothesis was that delinquency and continuation in delinquent behavior is a consequence of (1) the ineffectiveness of control structures and techniques of social

groups and/or (2) the relative absence of ineffectiveness of previously established con- trol structures and techniques of the person in producing and enforcing behavior which is in conformity with the norms of the

social system to which legal penalties are attached. Our observations show (1) that delinquent recidivists are less often than non-recidivists members of social groups and

live in a social milieu which is characterized

by norms and effective techniques in pro- ducing conformity behavior contra delin- quency, (2) that delinquent recidivists less often accept or submit to the control of

social groups which enforce such conformity behavior than do non-recidivists, and (3) that delinquent recidivists are less often persons with mature ego ideals or non- delinquent social roles and appropriate and flexible rational controls which permit the individual to guide action in accord with non-delinquent group expectations.

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

FAILURE OF PERSONAL AND SOCIAL CONTROLS 205

SELECTION OF THE PREDICTORS

The question may then be raised: Do these measures of personal and social con- trol provide valid prediction of delinquent recidivism? The evaluation of these measures as predictors can be undertaken in terms of the theory of prediction. The central problem of the theory of prediction is to make the best prediction for each case by minimizing the total error of prediction for the population as a whole on the principle of maximum probability. The best predic- tion is obtained when the accuracy or effi- ciency of a prediction instrument is at a maximum.

A central problem in maximizing accuracy and efficiency of a prediction instrument is to select items as predictors which make for valid prediction. The discriminative ability of a prediction instrument is in large part a function of the discriminative ability of the predictors. Two types of predictors selected by an item analysis may be differen- tiated on the basis of their discriminative ability: (1) Accurate predictors which do not yield predictive efficiency when predic- tion is made from the factor, (2) Efficient predictors which yield predictive efficiency when prediction is made from the factor. In the long run, efficient predictors yield more accurate and efficient prediction than do accurate predictors.17 Of the factors ini- tially selected in terms of the theory of personal and social controls, four were classi- fied as efficient predictors according to the theory of prediction. The remaining factors were accurate predictors.

Examining the four efficient predictors with a relatively high association with the criterion in Table 4, we observe that pre- diction from the diagnosis and recommenda- tions for treatment made by court workers and psychiatrists yields the most efficient prediction. Prediction from truancy or de- portment record, however, yields relatively low predictive efficiency. When predictions are made from these four efficient predictors

17 See Albert J. Reiss, Jr. "The Accuracy, Effi- ciency and Validity of a Prediction Instrument," op. cit., Chapters VIII and IX.

(Prediction Device I in Table 4) valid pre- diction is obtained, i.e. efficient prediction in expectancy and validation samples. Pre- diction from the diagnosis and recommenda- tions for treatment made by court workers and psychiatrists, however, is only slightly less efficient than prediction from the battery of four efficient predictors. Further we ob- serve that prediction from the three efficient predictors-adequacy of personal controls, truancy and deportment records (Prediction Device II in Table 4)-yields only slightly less efficient prediction than prediction from diagnosis and recommendations for treatment.

Viewing the prediction batteries in terms of the theory of personal and social controls, we note that prediction of court workers and psychiatrists are based on the evalua- tion of adequacy of personal controls with respect to acceptance of or submission to social controls. Three major categories of prediction are distinguished: (1) relatively strong personal controls; delinquent should succeed in his home community; (2) rela- tively inadequate personal controls; delin- quent may succeed in an environment of relatively strong social controls while under intensive supervision; (3) marked inade- quacy of personal controls; delinquent will not accept or submit to social control with- out psychotherapy or treatment in a com- munity or institutional setting. Similarly, prediction from strength of personal con- trols, truancy and deportment records pro- vide measures of the adequacy of the delinquent's personal controls and the ex- tent of the delinquent's acceptance of or submission to social control.

From the remaining factors selected in terms of the theory of personal and social controls, four of the factors classified as accurate predictors with a relatively high association with the criterion were selected and a prediction device constructed. Pre- diction from the accurate predictors did not yield valid prediction (Prediction Device III in Table 4). When predictions were made from the four accurate predictors for a follow-up sample of cases, negative predic- tive efficiency was obtained. This means

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

206 AMERICAN SOCIOLOGICAL REVIEW

that more errors in prediction are made in prediction from the battery of four accurate predictors than would be made in prediction from the over-all violation rate. Viewing the four accurate predictors in terms of the theory of personal and social controls, we

TABLE 4. EFFICIENCY OF PREDICTION DEVICES IN EXPECTANCY AND VALIDATION SAMPLES

Expectancy Sample Validation Sample

(N=736) (N=374)

Prediction Devices1 Prediction Predictive Prediction Predictive Errors' Efficiency' Errors' Efficiency'

Prediction Device I ....... ...... 212 +12.0 109 +17.4 (1) Recommendations for treat- ment .................... 215 +10.8 112 +15.1

(2) Adequacy of personal con- trols ..................... 231 + 4.2 121 + 8.3

(3) Truancy record ........... 231 + 4.2 130 + 1.5 (4) Deportment record ....... 239 + 1.3 131 + 0.8

Prediction Device II. (Predictors 2, 3 and 4) ........ 209 +13.2 118 +10.6

Prediction Device III ........... 236 + 2.1 139 - 5.3 (5) Residence in neighorhood area ..................... 241 0.0 132 0.0

(6) Residential mobility ....... 241 0.0 132 0.0 (7) Marital relationships of nat- ural parents ........ ...... 241 0.0 132 0.0

(8) Moral ideals and/or tech- niques of control by parents 241 0.0 132 0.0

Prediction Device IV (Predictors 1 to 8 inc.) ........ 214 +11.2 120 + 9.1

1 The total score for each case is obtained by scoring significant, unfavorable, efficient subcategories of a factor. Cf. Tables 1, 2 and 3 for significant subcategories of the factors.

Prediction from average violation rate results in 241 errors in the expectancy sample. Prediction from average violation rate results in 132 errors in the validation sample. 'The percent reduction in the error of prediction is employed as the measure of relative predictive effi-

ciency. It is here defined as the ratio of the difference between errors in prediction from total violation rate and score groups of a prediction instrument to the error or prediction from total violation rate.

observe that they are measures of the strength of social controls. Predictors em- ployed in this research as measures of family and community controls did not yield valid prediction of delinquent recidivism.18

18 The addition of any of the items which are measures of family and community controls to the battery of four accurate predictors does not make for valid prediction.

The results suggest that efficient predic- tion of delinquent recidivism is obtained when we use items as predictors which are measures of the adequacy of personal con- trols of the individual and his relation to social controls in terms of the acceptance

of or submission to social control. Such measures of personal and social control ap- pear to yield more efficient prediction of delinquent recidivism than items which are measures of the strength of social control. In the present research, prediction from items which are measures of the adequacy of personal controls and of the acceptance of or submission to social controls was 22

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms

DISCUSSION 207

per cent more efficient than prediction from items which are measures of the strength of social controls. From the standpoint of the theories of prediction and delinquent recidivism, the question may, however, be raised whether the combined set of pre- dictors yields more efficient prediction than prediction from either set of predictors. A prediction device using the entire set of items was therefore constructed. Our results show that prediction from the combined set of predictors yields less valid prediction than prediction from items which are measures of the adequacy of personal controls and

the acceptance of or submission to social control (compare Prediction Devices I and IV in Table 4).

In predicting delinquent recidivism, then, research might well be directed toward the isolation of items which are measures of

personal control re non-delinquent behavior

and items which are measures of the accept- ance of or submission to institutions which

exercise control contra delinquent behavior. Such measures of personal and social control may be expected to yield valid prediction of delinquent recidivism.

DISCUSSION

JOSEPH N. SYMONS

Utah State Agricultural College

Since 1928, when Dr. E. W. Burgess published the results of an investigation of the possibility of parole prediction, a large number of studies in the field have appeared. Among names to be remembered are: Sheldon and Eleanor Glueck, Walter Argow, William F. Lanne, Ferris F. Laune, Elio D. Monachesi (see his article in the American Sociological Review for February,

1945), Clark Tibbits, Coufrtland Van Vechten, George B. Vold, Ashley Weeks, Lewis M. Ter- man and others. These studies have shown:

a. How specific personal and social character- istics of individuals can be used to predict their conduct when they are placed in designated situations.

b. That the social scientist can definitely move toward more objectivity.

c. That "practical men," unfriendly to the theoretical, need not feel that nothing of value can come to them from the theoreti- cal approach.

In spite of gains from such studies, practical men have not as yet used the new tools freely and willingly, nor have social scientists done ample testing of the usefulness of the newly created prognostic instruments.

As we view the Reiss paper as a whole, we should commend him for:

a. Moving forward to test the usefulness of prognostic instruments.

b. His search for good predictors of delin- quent behavior. Permit me to read again his purpose, as stated early in his paper. It reads: "The purpose of the present paper is to isolate a set of personal and social controls which are associated with delinquent recidivism and to evaluate them as prognostic of recidivism."

c. Also his seeking a few good predictors. Some of the earlier sets were large and un- wieldly, and because of their size, were discouraging to practical men from the standpoint of their use.

d. His laudable inroads into the use, as pre- dictors, of personal and social controls. Attention to primary group controls, com- munity and institutional controls and per- sonal controls, shows a grasp of the kinds of control which really count in human behavior.

Along with the strengths of the Reiss paper came those items which I am sure some people in the audience would choose to question, or discuss further. First, I feel confident someone here, as he read the title of the paper-"De- linquency as the Failure of Personal and Social Controls"-asked silently, "Is this another sin- gle-cause explanation of delinquency and crime?" The author did not mean it to be such, I am sure. His concern has been the seeking out of predictors and the testing of them. Also, we

This content downloaded from �������������152.46.28.191 on Tue, 03 Mar 2026 15:19:49 UTC�������������

All use subject to https://about.jstor.org/terms