Similarity between deviant peers: Developmental trends in influence and selection*

• Author: Daniel T. Ragan
• Published date: 01 May 2020
• Date: 01 May 2020
• DOI: http://doi.org/10.1111/1745-9125.12238

Received: 21 April 2017 Revised: 16 October 2019 Accepted: 2 December 2019

DOI: 10.1111/1745-9125.12238

ARTICLE

Similarity between deviant peers: Developmental

trends in influence and selection*

Daniel T. Ragan

University of New Mexico

Correspondence

DanielT. Ragan, Department of Sociology,

Universityof New Mexico, MSC05 3080, 1915

RomaNE Ste. 1103, Albuquerque, NM 87131.

Email:dragan@unm.edu

Fundinginformation

NationalInstitute on Drug Abuse, Grant/Award

Numbers:R01-DA018225, RO1-DA013709;

NationalInstitute of Child Health and Develop-

ment,Grant/Award Number: R24-HD041025;

NationalInstitute on Alcohol Abuse and Alco-

holism,Grant/Award Number: AA14702; W.T.

GrantFoundation, Grant/Award Number: 8316

∗Grantsfrom the W.T. Grant Foundation

(8316),National Institute on Drug Abuse (R01-

DA018225),and National Institute of Child

Health and Development(R24-HD041025)

supported this research. The analysesused

data fromPROSPER, a project directed by R.

L.Spoth, funded by the National Institute on

Drug Abuse (RO1-DA013709)and the National

Instituteon Alcohol Abuse and Alcoholism

(AA14702).The content of t his article is solely

the responsibilityof the authors and does not

necessarilyrepresent the official views of the

NationalInstitutes of Health. The author thanks

WayneOsgood, Derek Kreager, Scott Gest,

DavidJohnson, David Schaefer, and Jacob

Youngfor helpful comments.

Abstract

The association between an adolescent’s ownbehavior and

that of his or her peers remains a key empirical finding

in the study of delinquency, and this similarity is often

explained in criminology by invoking processes of social

influence and homophily. Adolescence is a period of

rapid change for both individuals and their surroundings,

however, and influence and homophily are often discussed

without attending to their development over time. In the

current study, I employ longitudinal social network models

to estimate social influence and homophily related to

alcohol and cigarette use and to determine whether there

is change in the strength of these processes. Furthermore,

I test whether the broader social environment conditions

these processes and their observed changes. The results

from this study indicate that although influence from peers

decreases from early to mid-adolescence, homophily on

drinking and smoking increases. There is also evidence

that school-level opportunities and expectations affect

social influence and homophily but do not account for the

average changes observed in these processes.

KEYWORDS

adolescence, delinquent peers, homophily, peer influence, social networks

Adolescence represents a unique stage in the life course for changes in both social relationships and

drug and alcohol use. Friends often share similar characteristics, and associating with deviant peers is

one of the most consistent predictors of whether an adolescent engages in deviant behavior.At the same

time, adolescence is also a period during which rates of drug and alcohol use rise dramatically (John-

ston, O’Malley, Miech,Bachman, & Schulenberg, 2017). The convergence of these trends—similarity

336 © 2020 American Society of Criminology wileyonlinelibrary.com/journal/crim Criminology.2020;58:336–369.

RAGAN 337

between peers and increases in drug use—has led to peers being identified as a particularly salient fac-

tor in the development of drug use (Hawkins, Catalano, & Miller, 1992; Oetting & Beauvais, 1987).

But although the tendency to associate with others who share similar attributes is well established for

drinking and smoking in adolescence, the processes that produce this similarity are less understood. In

particular, questions of how, or whether,these processes change during adolescence have gone mainly

unanswered despite this period being marked by notable changes in both deviant behaviors and peer

relations.

In traditional criminological theory, covariates of offending during adolescence, including deviant

peers, are often explained without considering how these variables evolve within this stage of the life

course. In the current study, I test how two types of processes that produce similarity among friends—

influence and homophily—change in strength during adolescence in regard to alcohol and cigarette

use, and I examine whether these changes are conditioned by other changes in the broader social con-

text that occur during this time. I begin by reviewing the criminological literature on delinquent peers

and how processes of influence and selection contribute to similarity between delinquent peers. Next, I

turn to a discussion of why criminologists should expect change in these processes during adolescence

and examine the empirical research on this topic. I also discuss why changes in friendship influence

and selection processes may be linked to the changes in the broader social environment that occur

concurrently during adolescence. After describing my data and methodology, I present evidence that

although influence from friends decreases during adolescence, the role of selection in producing sim-

ilarity increases during this period. Additionally, network-wide levelsof oppor tunity and expectations

help predict the strength of these social processes, but they do not explain the average changes in the

processes that occur.

1SIMILARITY: THEORY AND EVIDENCE OF SELECTION

AND INFLUENCE

One of the most widely observed social phenomena is that individuals tend to associate with others

who possess similar traits and characteristics as themselves (Brechwald& Prinstein, 2011; McPherson,

Smith-Lovin, & Cook, 2001). In the study of adolescent deviance, this observation is one of the most

consistent empirical conclusions available in the literature: Adolescents who engage in deviant behav-

iors are more likely to associate with peers who are also deviant (Warr, 2002). This finding extends

to behaviors such as alcohol and tobacco use, and explaining why deviant adolescents associate with

one another remains a key theoretical question whose answer has implications for both theory and the

development of prevention programs.

Many different processes may produce similarity between peers (McPherson et al., 2001; Shalizi

& Thomas, 2011), but two in particular are often at the forefront of understanding peer delinquency:

influence and selection. Peer influence, also known as “peer socialization,” occurs when individu-

als adopt the beliefs or behaviors of those with whom they associate. Sutherland’s (1947) theory of

differential association, for example, centers on the idea that having delinquent friends brings about

increased exposure to beliefs that promote deviance. Social learning theory (Akers, 1985; Burgess &

Akers, 1966) builds on this proposition by noting that those with delinquent peers will also be subject

to increased social reinforcement of deviance and will haveadditional opportunities to model deviance.

From a peer influence perspective, delinquent friends thus represent a cause of delinquent behavior.

Homophilic selection—hereafter “homophily”—is the “tendency for friendships to form between those

who are alike in some designated respect” (Lazarsfeld & Merton, 1954, p. 23) and also leads to sim-

ilarity among friends. In this case, delinquent individuals seek out people like themselves as friends.

338 RAGAN

As a result, delinquency not only precedes the association between those who are similar but also may

contribute to why the similarity exists.

Although once presented as competing mechanisms, a long line of research into delinquent peers has

found support for both influence and selection processes, indicating that each contributes to similar-

ity among friends (Cohen, 1977; Kandel, 1978; Osgood, Feinberg, & Ragan, 2015). That is, although

influence was once cited as the underlying mechanism responsible forpeer similar ity byproponents of

learning theories of deviance (Akers, 1985; Burgess & Akers, 1966; Sutherland, 1947), and homophily

was invokedto explain peer similarity in control theor ies (Gottfredson& Hirschi, 1990; Hirschi, 1969)

and by other approaches (e.g., Glueck & Glueck, 1950), current research on delinquent peers generally

recognizes the presence of both processes. Moreover, the existence of both processes implies that they

are complementary, rather than competing, explanations for similarity: Friendship selection shapes

which individuals have the opportunity to exert influence (Oetting & Beauvais, 1986; see also Dishion

& Owen, 2002; Osgood et al., 2013). Similarly, in Thornberry’s (1987) interactional theory of delin-

quency, the relationship between delinquent peers and delinquent behavior is reciprocal: Associating

with delinquent peers leads to delinquent behavior, which in turn leads to associating with additional

delinquent peers, who continue to reinforce delinquency.

Understanding why deviant adolescents associate with one another is aided by recent advances in

longitudinal social network analysis that offer the ability to model processes of both selection and

influence simultaneously. One examplecan be found in the results from stochastic actor-based models

estimated by the SIENA software developedby Snijders and colleagues (Snijders, 2001, 2005; Steglich,

Snijders, & Pearson, 2010). Evidence is provided in this research for both influence and selection

on several behaviors, including drinking and smoking, in diverse samples (Knecht, Burk, Weesie, &

Steglich, 2011; Osgood et al., 2015; Schaefer, Haas, & Bishop, 2012; Steglich, Sinclair, Holliday, &

Moore, 2012; see also Gallupe, McLevey, & Brown, 2019, for a review of studies on other types of

offending).

2CHANGES IN INFLUENCE AND SELECTION:

THEORETICAL CONSIDERATIONS

Similarity between delinquent peers increases in adolescence (Osgood, Ragan, Dole, & Kreager,2019),

and this observation occurs against the backdrop of adolescence itself, a unique stage in the life

course marked by simultaneous development across multiple domains. In a review of criminologi-

cal research on patterns of crime over the life course, Sampson and Laub (1992, pp. 64–65) made a

series of recommendations for advancing this work—such as employing longitudinal data and giv-

ing attention to transitions occurring during adulthood—and argued that “change must be consid-

ered part of the explanatory framework in criminology.” These recommendations arise at least in

part as a result of Sampson and Laub’s (1992, p. 64) observation that, “The age-crime curve has

had a profound impact on the organization and content of sociological studies of crime by channel-

ing research to a focus on adolescents.” Paradoxically, however, the typical criminological approach

to studying adolescence has been limited in ways that yield an incomplete understanding of behav-

ior for this same reason, in that it overlooks the dramatic developmental changes in this period that

bridges childhood to adulthood (but see Thornberry, 1987). That is, adolescence is a period marked by

rapid changes, and in empirical research on criminological theory, scholars have often failed to incor-

porate these changes into explanations of deviant behaviors. These changes, which occur in both the

individual and the larger social environment, may be consequential for peer influence and selection

processes.