Religion and Delinquency:A Systematic Review of the Literature

Coordinating Council for Juvenile Justice and Delinquency

PreventionWashington, DC

December 5, 2008

Byron R. JohnsonProfessor of Sociology

Director, Institute for Studies of ReligionDirector, Program on Prosocial Behavior

Baylor UniversityWaco, Texas

Religion & Health Outcomes

• Hypertension• Mortality• Depression • Suicide• Promiscuous sexual behavior• Drug & alcohol use• Crime & Delinquency

Figure 1. Research Examining the Relationship between Religion and Health Outcomes (2002)

22

6

1

46

14

1

70

29

4

61

9

0

37

10

84

10

3

50

3

1

37

8

1

01020304050

60708090

100

Hypertension Mortality Depression Suicide Sexual Behavior Alcohol Use Drug Use Delinquency

Figure 1. Research Examining the Relationship between Religion and Health OutcomesEight Fields of Study (total of 498 studies reviewed)

Beneficial OutcomesNA/Mixed OutcomesHarmful Outcomes

%

Figure 2. Research Examining the Relationship between Religion & Health Outcomes (2002-2008)

31

6

1

57

15

1

116

43

4

100

19

3

101

40

110

113

105

81

134

19

2

0%

20%

40%

60%

80%

100%

Hypertension Mortality Depression Suicide Sexual Behavior Alcohol Use Drug Use Delinquency

Eight Fields of Study (total of 894 studies reviewed)

Beneficial OutcomesNA/Mixed OutcomesHarmful Outcomes

%

Religion & Prosocial Outcomes

• Well-Being• Hope/Purpose/Meaning in Life• Self-Esteem• Educational Attainment

Figure 3. Research Examining the Relationship between Religion Prosocial Outcomes (2002)

80

18

1

25

5

0

15

8

1

16

21

0102030405060708090

%

Well-Being Hope Self Esteem Educational

Attainment

Four Fields of Study ( total of 171 studies reviewed)

Beneficial OutcomesNA/Mixed OutcomesHarmful Outcomes

Figure 4. Research Examining the Relationship between Religion & Prosocial Outcomes (2002-08)

104

23

1

25

5

0

27

12

1

16

21

0%

20%

40%

60%

80%

100%

%

Well-Being Hope Self Esteem Educational

Attainment

Four Fields of Study ( total of 217 studies reviewed)

Beneficial OutcomesNA/Mixed OutcomesHarmful Outcomes

Meta-Analysis of Research on the Impact of Religion on Delinquency

Total of 263 studies reviewed

Studying Organic ReligionReview Summary

• This review provides overwhelming evidence that higher levels of religious involvement and practices make for an important protective factor that buffers or insulates individuals from deleterious outcomes.

• This review of studies on organic religion documents that religious commitment or practices make for an important factor promoting an array of pro-social behaviors and thus enhancing various beneficial outcomes.

A Study of the Cumulative Advantage of Religiosity

Sung Joon Jang and Byron R. JohnsonDepartment of Sociology

Institute for Studies of Religion (ISR)Baylor University

• Does early involvement in religion matter?– Pre-teen involvement in religious activities– Religious involvement after childhood– Drug use during adolescence– At-risk drinking/smoking in young adulthood

• What does scientific research say about long-term effects of religious involvement?

• The more religious involvement, the less likely to use drugs, licit and illicit

• How so?– Increased control– Decreased deviant learning– Decreased strain & distress– Enhanced positive self-concept– Perhaps uniquely religious influence?

• Cumulative disadvantage in criminology– Delinquency Negative structural

consequences (e.g., labeling, negative life events, etc.) Crime

– A negative causal loop

• Cumulative advantage of prosocial behavior– Religious involvement Positive consequences

(e.g., “social capital”) Prosocial behaviors– Does it work?

• It seems to (Jang, Bader, & Johnson, 2008), but …

Explaining the Positive Consequences of Religiosity

• Religiosity Protective factors– Social control theory

• Attachment to parents, commitment to education, etc.

• Religiosity Risk factors– Social learning theory

• Association with drug-using peers– Social strain theory

• Negative life events– Symbolic interactionist theory

• Reflected appraisal of self as “trouble maker”

Data

• National Education Longitudinal Study of 1988 (NELS:88)– Nationally representative sample (n = 12, 144)– Four follow-ups (1990, 1992, 1994, and 2000)

• Time 1 (Wave 1): 8th grade, average age = 14.4• Time 2 (Wave 2): 10th grade, average age = 16.4• Time 3 (Wave 3): 12th grade, average age = 18.4• Time 4 (Wave 5): 8 yrs after high sch, average age

= 26.4

– Data analysis in two parts

Theoretical Model of Drug Use

Religiosity

Protectivefactors

Druguse

Riskfactors

Religiosity

Protectivefactors

Druguse

Riskfactors

Religiosity

Protectivefactors

Druguse

Riskfactors

__

Sociodemographiccontrol variables:

SexAgeRace

Family SESIntact family

Religiousdenomination

+

+

+

_

_

_

+

+

_

+_

__

+

+

+

+

_

+

_

+

_

+

Time1 (1988)8th grade

Ages 13-16 (14.4)

__

+

_

_

Religiosity

Bingedrinking& Heavysmoking

+

+

+

+

Time 2 (1990)10th grade

Ages 15-19 (16.4)

Time 3 (1992)12th grade

Ages 17-20 (18.4)

Time 4 (2000)Ages 25-28 (26.4)

_ _

_

+

+

+

+

+

+

+

_

_+

__

+

+

+

_

_

__

+

_

Pre- orearly teenreligious

involvement

+

_

_

_+

+

+

_

__

Estimated Model of Drug Use

Religiosity

Protectivefactors

Smoking

Religiosity

Protectivefactors

Druguse

Riskfactors

Religiosity

Protectivefactors

Druguse

Riskfactors

Time1 (1988)8th grade

Ages 13-16 (14.4)

Time 2 (1990)10th grade

Ages 15-19 (16.4)

Time 3 (1992)12th grade

Ages 17-20 (18.4)

Pre-teenreligiosity Risk

factor

. 39

4

.139

.248

.109

.062

.621

.034

-.0

86

-.1

71

.299

-.1

6 3

.109

.201.1

31

-.2

61

.197

-. 05

2

.024

.455

-.3

15

.337

-.0 5

5

-.0

72

.17

1-.

112

-.1

23-.021

-.036+

-.01 8

-.038

.025+

-.034

.387

.126

.149

.243.0

19+

-.094

.137 .216

.040

-.014

.018

+

.025

.22

9-.040+

-.060

Findings

• Early involvement in religion during pre-teen years tends to decrease:– use of licit & illicit drugs during adolescence– binge drinking & heavy smoking in young

adulthood

• Empirical evidence of cumulative advantage of religiosity as pre-teen religiosity increases protective factors and decreases risk factors of drug use.

Summary and Conclusions

Early involvement in religion (i.e., pre-teen religiosity) tends to have long-term, prosocial influence on an individual’s life.

Prosocial influence of religion is developmental in that it tends to be not only direct but also indirect (specifically, increasing other prosocial factors and decreasing antisocial factors).

Religiosity and drug use tends to have reciprocal (i.e., two-way) relationships over time.

A key question for future research: Would religious involvement, whether sudden or gradual,

be able to break the cycle of drug use and its risk factors?

Figure 1. Structural Equation Model of Youth’s Drug Use during Adolescence & Young Adulthood (n = 1,377)* p < .05 (one-tailed test), + p < .05 (two-tailed test), f = fixed coefficient, smc = squared multiple correlation (i.e., R2)

Race(black)

Parent'sreligiosity T1

smc = .084

Deviantpeers T2

smc = .200

Youth'sreligiosity T2

D1

D2

D4

D6Parent'sdrug use T1

alcohol

e

tabacco

.509*

illegal drugs

smc = .335Pro-drugpeers T3

illegal drugs 1

smc = .610Youth'sdrug use

T3

illegal drugstabaccoalcohol

smc = .378

Youth'sreligiosity

T3

activities

pray

salience

smc = .424Youth'sdrug use

T2

alcohol tabacco illegal drugs

alcohol 1

D5

Scriptures

services

D3

e e

e e e

e

e

e

e

e

e

ee

e

e e e

.559*

alcohol 2

illegal drugs 2

.107*

-.208*

-.130*

-.108* -.119*

.106* .307* .191*

GenderAge

Family size T1Family SES T1

Family disruption T1Residential mobility T1

RegionAttachment to parent T1

Self-control T1Negative emotions T1

-.019.124*

.036

.187*

.052

.133*

-.063

Religion, Prosocial Learning, Self Control, and Delinquency

Scott A. DesmondPurdue UniversityJeffery T. Ulmer

Penn State UniversityChristopher D. Bader

Baylor University

Why Might Religiosity Inhibit Delinquency?

• Social Learning: Religious observance and taking religious teaching seriously involves learning prosocial norms and values.

• Self Control: Religious observance may build self control.

This Study’s Goals• We examine the degree to which religiosity,

and membership in particular religious traditions, increases self control.

• We assess the effects of religiosity, religious denominations, and self control on delinquency, and the extent to which religiosity has an indirect effect on delinquency through self control.

• Finally, we investigate whether the effect of self control on delinquency varies depending on adolescents’ religiosity.

• We use data from the National Longitudinal Survey of Adolescent Health (Add Health).

Prior Theory and Research

• Concern with self control has a long history in criminology, but very few studies simultaneously examine religiosity, self control, and crime/delinquency.

• The consensus of prior research indicates that religiosity has a moderate but consistent effect on delinquency, particularly in stronger “moral communities,” and particularly for drug/alcohol use.

• Research has shown that religiosity’s effects are partially, but not fully, indirect, through social learning and conventional social bonds.

What else can help explain religiosity’s ability to insulate youths from delinquency?

•Self control is a good candidate. Every major religion tries to promote self control in the face of behavior seen as wrong.

•A number of studies: a) endorse the concept of self control, b) suggest that it is malleable after late childhood, and c) suggest that factors other than the parenting practices identified by Gottfredson and Hirschi (1990) affect it.

•“Self control is itself a product of social learning, and therefore any variation that it accounts for ultimately depends on social learning.” –Ronald Akers (1998, p. 162).

• Religious socialization and exposure to religious activities would seem to be a potentially important social learning process by which self control could be developed and increased. Religion should thus increase self control.

New Developments in Thinking about Self Control

• Psychologists have developed a “muscle” or “strength” model of self control– Self control is a cognitive resource that

is temporarily depleted whenever it is exercised, just as a muscle is temporarily fatigued when it is used. This model also implies that self control should grow stronger with regular “exercise.”

Religion and Self Control

•“Religious organizations, as an external source of discipline, can be very helpful to people’s personal self control endeavors.” Geyer and Baumeister (2005, p. 418).

•Religion can:– Provide motivation for self control– Foster self monitoring– Help people manage desires– Help people cope with emotional distress

(which can trigger self control failures)

Does the Effect of Self Control on Delinquency Differ by Degree of Religiosity?

• Self control might have a stronger negative effect on delinquency among the more religious, since it is in situations where one must choose whether to engage in a deviant act, in the face of one’s own moral beliefs, that self control must be mobilized. No prohibiting beliefs, no need for self control. (Wikstrom and Treiber 2007)

Religiosity and the Dynamics of Delinquency

• Jeffery T. Ulmer• Penn State

University

• Sung Joon Jang• Baylor University

• Christopher D. Bader

• Baylor University

• Scott A. Desmond

• Purdue University

• Byron R. Johnson• Baylor

University

This Study’s Goals• We examine the degree to which religiosity

discourages the initiation and persistence of delinquency, as indicated by marijuana use.

• We assess the degree to which any effects of religiosity on patterns of marijuana use are meditated by (and therefore attributable to) non-religious factors such as peers, parental and school attachment, self control, and strain/negative emotionality.

• We use data from the first three waves of the National Longitudinal Survey of Adolescent Health (Add Health).

Why would religiosity discourage initiation and persistence in marijuana use? Why would it encourage desistence even if someone begins using?

• Religiosity might discourage friendships with marijuana using friends (or the influence of such friends on one’s own behavior).

• Religiosity might foster prosocial, conventional attachments and activities.

• Religiosity might foster self control (a la Desmond, Ulmer, and Bader)

• Religiosity might buffer individuals from strain and negative emotionality, and/or discourage coping with strain/negative emotions with drug use.

• Religion might exert prosocial influences in its own right, influences that are not attributable to non-religious social processes (Pargament et al. 2005).

• Religious experience and practice might thus exert unique influences discouraging forms of delinquency like marijuana use.

Prior Research: Key studies• In general, little research on religiosity

and desistence delinquency/drug use.

•Doris Chu (2007): Using NYS data, she found that frequency of church attendance was associated with desistence from marijuana use and other drugs, and perceived religious importance was negatively related to initiation of drug use. • Drawback: limited data, few controls for other

important known predictors of drug use.

• Peggy Giordano and associates’ studies (2008; 2007; 2002) have found qualitative evidence that religiosity can be a “hook for change” and a source of “prosocial capital” that fosters desistence from crime among a sample of highly socioeconomically disadvantaged offenders.

• However, their quantitative analysis found that religiosity, by itself, did not have a significant effect on desistence from crime, when controlling for other criminogenic factors.

• Other suggestive evidence: Studies show that faith based approaches help prevent recidivism among adult prisoners (Johnson and Larson 2003; Johnson 2002; Hercik 2004).

Data

• National Longitudinal Study of Adolescent Health (Add Health), Waves 1-3.– Stratified sample of 80 high schools, 52

middle schools.– Over 8,000 respondents available for our

Wave 1-3 analyses.– Our analyses correct for unequal probability

of selection and clustering of students in schools.

Method

• Multinomial Logistic Regression

– Used for analyses of multiple outcome categorical data.

– Allows us to examine the likelihood of respondents exhibiting one of the patterns of marijuana use we are interested in (initiation, intermittent, desistence, persistence) versus the others.

Dependent Variable: Patterns of Marijuana Use Over Time

• Never used• Initiation=not using at Wave 1 but using at

Wave 2 and/or 3, • Intermittent=not using at Wave 1, using at

Wave 2, not using at Wave 3, or vice versa. • Desistence=using at Wave 1 but stopping

at Wave 2 and not using at Wave 3, or, using at Wave 1 and 2 and stopping at Wave 3

• Persistence=using at all three times.

Religious Variables

• Religious tradition (Catholic, evangelical protestant, black protestant, mainline protestant, other, none)

• Parent’s religiosity• Biblical literalism• Born again Christian• Adolescent’s religiosity (church

attendance, importance of religion, frequency of prayer) 

Control Variables

• Sex• Age• Race/ethnicity • Living with both biological parents• On welfare (AFDC/TANF, SSI, food

stamps, housing subsidy)• Parents’ education• Grade point average

Mediating Variables

• Parental attachment• School attachment• Number of “best friends” using

marijuana at least once a month• Self control scale• Negative emotionality scale

Highlights of Results

• Religiosity strongly discourages persistent marijuana use and encourages desistence.

• Religiosity may tend to discourage initiation of marijuana use.

• Some of religiosity’s effects are mediated by other known predictors of delinquency, religiosity’s effects remain significant and strong even when controlling for the effects of other variables.

Conclusions

• Religiosity discourages persistent marijuana use, and may discourage initiation of marijuana use.

• Religiosity does so partly because of its indirect influence on friendships, prosocial attachments, self control, and coping with strain/negative emotionality.

• A strong part of religiosity’s effects are unique, and not due to the above non-religious factors.

www.ISReligion.org