Revisiting labeling theory: empirical test of informal labeling process

ABSTRACT

Despite the revitalization of labeling theory in explaining deviant behavior, the theoretical validity of the proposition regarding the informal labeling process has been largely neglected. To examine the proposed informal labeling process within the labeling theory, longitudinal path analyses were conducted using structural equation modeling, drawing on a nationally representative sample of adolescents in South Korea. The results revealed that engagement in delinquent behavior at Wave 1 had a significant impact on how others informally labeled the individual at Wave 2. Also, informal labeling by others at Wave 2 significantly influenced the development of deviant self-identity at Wave 3, and this deviant self-identity increased the likelihood of engaging in subsequent delinquent behavior at Wave 4. Namely, the hypothesized informal labeling process in labeling theory was supported: informal labeling based on previous delinquent behavior → development of deviant self-identity → engaging in subsequent delinquent behavior among adolescents. The findings from this study highlight the theoretical importance of the informal labeling process in explaining delinquent behavior among adolescents from a longitudinal perspective. The research implications are discussed.

KEYWORDS:

• Labeling theory
• informal labeling
• deviant self-identity
• delinquency

Introduction

Labeling theory posits that adolescents who are sanctioned and identified as deviant may interpret their stigmatization as a master status, thereby altering their self-identity and subsequently adjusting their behavior (Ascani 2012; Paternoster and Iovanni 1989; Triplett and Upton 2015). Put differently, labeling theory assumes that perceived deviant societal reactions from others contribute to the development of a deviant self-identity, which then leads labeled individuals to engage in subsequent deviant behavior (Adams et al. 2003; Becker 1963; Krohn and Lopes 2015; Lemert 1951, 1967). Labeling theory posits two types of labeling: 1) Formal labeling and 2) Informal labeling. According to labeling theory (Bernburg 2019; Triplett and Upton 2015), formal labeling refers to the reactions by official institutions within the criminal justice system, such as arrest and conviction. On the other hand, informal labeling pertains to the reactions of parents, friends, and teachers (Triplett and Jarjoura 1994). While early labeling theorists primarily focused on formal labeling and the formal labeling process, recent studies on the association between labeling and subsequent deviant behavior have highlighted a strong impact of informal labeling on adolescents’ deviant behaviors (Kavish, Mullins, and Soto 2016; Kim and Park 2022). However, questions regarding the theoretical and empirical validity of informal labeling still persist (Restivo and Lanier 2015). For instance, existing studies have primarily concentrated on examining the impact of informal labeling rather than exploring the dynamic process proposed by labeling theory (Kaplan and Lin 2005; Liu and Bachman 2021). Namely, researchers have rarely investigated the existence of informal labeling processes, which are crucial for comprehending the transition from deviant labeling to subsequent deviant behavior. As a result, the central proposition of labeling theory regarding informal labeling has not been comprehensively tested (Bernburg 2019). In this line, the current study is designed to address an important but relatively neglected and understudied issue: the informal labeling process and delinquency (Zhang 1997). To achieve this goal, the study presents the propositions on the informal labeling process in labeling theory and reviews the limited research that has examined these issues. Subsequently, an investigation into the reliability of a theoretical model of the informal labeling process is conducted using a longitudinal nationally representative sample of South Korean adolescents. It is noteworthy that previous studies on informal labeling have predominantly utilized Western samples (Brownfield and Thompson 2008; Hayes 2010; Kavish, Mullins, and Soto 2016; Matsueda 1992; Walters 2016), and the cultural differences between South Korean and Western society offer a valuable context for expanding upon previous research (Chen 2002). Therefore, this study aims not only to investigate the mechanism of informal labeling rigorously but also to examine its generalizability.

Theoretical rationale

Labeling theory

Becker (1963) and Lemert (1951) employed labeling theory to explain how the development of a deviant identity is associated with the continuation of deviant behavior. According to the labeling perspective (Becker 1963; Lemert 1951; Triplett and Jarjoura 1994), the initial or primary deviant behavior is considered unimportant since all individuals engage in deviance at some point. However, once individuals have been labeled by social control agencies or defined as deviants due to their primary deviant behavior, labeled individuals are more likely to encounter problems associated with the deviant label (Becker 1963; Bernburg and Krohn 2003; Krohn and Lopes 2015). When individuals develop a deviant self-identity as a result of being labeled by others, they are more likely to engage in further deviant behavior associated with that identity. Consequently, this behavior can trigger problematic social reactions (Ward and Tittle 1993). These problems, in turn, can increase the likelihood of continued engagement in deviance and criminal behavior throughout one’s life. Such a pattern may arise due to detachment from prosocial activities like school, alienation from conventional individuals, and changes in self-identity (Bernburg 2019; Lopes et al. 2012). For instance, labeled individuals may encounter social barriers that prevent them from enjoying the benefits of mainstream society, such as obtaining a diploma, establishing strong relationships with significant others, and participating in civic life (Ascani 2012). In this regard, the central proposition of labeling theory is that experiencing negative social labels such as ‘deviant,’ ‘delinquent,’ or ‘criminal’ contributes to future engagement in deviant behavior (Becker 1963; Lee et al. 2017; Lemert 1951).

Traditionally, labeling researchers have primarily focused on the role of official labels such as suspension or arrest in labeling studies, neglecting the symbolic nature of the labeling process that arises from informal labeling (Hayes 1997). However, informal labeling by parents, teachers, peers, or community members can result in exclusionary reactions toward youth and impact their self-concept (Bernburg 2019; Matsueda 1992). For example, from the framework of informal labeling, Lemert (1967) emphasized the role of deviant self-concept in motivating and controlling behavior, pointing out that individuals’ self-appraisal is shaped by their experiences with interaction with others. The outcome of this reflected appraisal is a self-fulfilling prophecy, further amplifying the process of deviance or secondary deviance (Bartusch and Matsueda 1996). That is, the evaluation and appraisal by others play a crucial role in influencing one’s behavior by shaping their self-identity. Consequently, adolescents who perceive themselves as deviant are more likely to engage in deviant behavior compared to those who view themselves as conformists (Bartusch and Matsueda 1996). In short, the implied informal labeling process can be illustrated as follows: Deviance → labeling → development of deviant identity → recidivism (Ward and Tittle 1993).

On the whole, labeling theory predicts that an adolescent’s experiences interacting with others who perceive them as deviant for their primary delinquent behavior (i.e., informal labeling) leads to the development of a deviant self-identity, which in turn leads them to engage in subsequent deviant behaviors (Bernburg 2019; Hayes 2010). Despite the importance of this robust theoretical foundation of the informal labeling process within labeling theory (Kavish, Mullins, and Soto 2016; Kubrin, Stucky, and Krohn 2009; Paternoster and Iovanni 1989; Triplett and Jarjoura 1994), the theoretical prediction regarding the mediating role of informal labeling by others and the development of a deviant self-identity between primary delinquency and subsequent delinquency has not been empirically evaluated.

Previous studies on informal labeling

There is evidence supporting the role of informal labeling through the significant influence of labels by others and deviant self-identity on involvement in delinquent behavior. For instance, Matsueda (1992) demonstrated that self-identity mediated the impact of parental labeling on subsequent delinquent behavior among male youths. Additionally, Bartusch and Matsueda (1996) discovered that youths’ self-perception as rule violators, influenced by reflected appraisals, had a substantial and statistically significant effect on their secondary delinquency, regardless of gender. Similarly, irrespective of gender, Brownfield and Thompson (2005) found that reflected appraisal by parents and peers, as well as self-concept, were significantly associated with delinquency even after controlling for social control factors such as attachment, commitment, and belief variables. These previous studies emphasized that the evaluation of informal labeling by significant others such as parents, peers, and teachers played a critical role in predicting adolescents’ future delinquent behavior by shaping their self-identity (Lopes et al. 2012). Also, regarding the proposed informal labeling process in labeling theory, Zhang (1997) conducted an empirical study to examine the propositions of the informal labeling process. Using a sample of 1,725 youth from two-wave data in NYS, he identified a significant association between informal labeling by others (i.e., parents, peers, and teachers) based on prior delinquency and adolescents’ perceptions of informal labeling. Furthermore, it was found that informal labeling by others resulted in isolation from school and home, which in turn increased the likelihood of engaging in subsequent delinquent behaviors. In a similar vein, Walters (2016) tested the hypothesis of the informal labeling process by investigating the relationship between reflected appraisals by others, deviant self-identity, and adult offending in a sample of 791 participants from the original 1942 and 1949 Racine cohorts. This study, although conducted on an outdated Western sample, demonstrated that informal labeling by others influences the development of deviant self-identity, thus supporting the proposition of the informal labeling process in labeling theory.

However, empirical examinations of the informal labeling perspective remain scarce (Bernburg 2019; Kavish, Mullins, and Soto 2016; Paternoster and Iovanni 1989). Furthermore, there are still questions about the generalizability of findings to other cultural contexts (Walters 2016). In other words, to date, very limited research has investigated the process of the formation of a deviant self-concept and its mediating role between informal labeling and subsequent delinquency, particularly in Western contexts (Bernburg 2019; Kubrin, Stucky, and Krohn 2009).

South Korean context

Rooted in Confucianism and collectivism, South Korean culture makes beneficial research settings for examining the process of informal labeling (Kim and Park 2022). In South Korean culture, adolescents are more exposed to others and more perceptive of others’ appraisals, which means they are more likely to be vulnerable to informal labeling by others in the establishment of identity (Lee 2018). As a result, considering those social forces, it might be appropriate to verify the process of informal labeling of South Korean adolescents. In this context, prior research employing South Korean samples has underscored the significant influence of informal labeling on deviant behavior, encompassing both status offenses and general delinquency (Kim and Park 2022; Lee 2018; Na and Paternoster 2019). Consistent with the predictions of labeling theory, these studies found that both adolescents’ deviant self-identity and informal labeling by others were associated with increased involvement in subsequent minor and serious delinquency. For instance, in a study with a nationally representative sample of 2,406 South Korean adolescents, Lee (2018) observed a direct effect of informal labeling on subsequent delinquency, even after controlling for primary indicators of delinquency such as association with delinquent peers, family attachment, and school commitment. In addition, Na and Paternoster (2019), in their longitudinal path analyses, found that the establishment of a deviant self-identity through appraisals from others was related to engagement in violent behavior of South Korean adolescents. These previous studies illustrate the significance of examining the theoretical framework of the informal labeling process to understand adolescents’ pathways to engage in delinquent behavior from a South Korean context.

The present study

As such, despite previous studies supporting the impact of informal labeling and deviant self-identity on subsequent delinquent behavior, the process of informal labeling remains largely unexplored (Lee, Menard, and Bouffard 2014; Triplett and Jarjoura 1994). Therefore, an attempt to verify of informal labeling process is imperative to examine the potential capacity of the labeling theory for explaining delinquency (Zhang 1997). To address this empirical research gap regarding the theoretical proposition of the informal labeling process, the current study aims to examine the relationships between primary delinquency, informal labeling, the development of deviant identity, and subsequent delinquency. Specifically, the study will investigate whether informal labeling based on primary delinquency influences the development of deviant self-identity and whether the relationship between informal labeling and subsequent delinquency is mediated by deviant self-identity in South Korean adolescents. To achieve these objectives, three main research questions will be addressed, and the hypothesized theoretical model is presented in Figure 1.

Figure 1. Hypothesized model.

Hypothesis 1:

Primary delinquency will be significantly related to informal labeling by others.

Hypothesis 2:

Informal labeling by others will be significantly associated with the development of deviant self-identity in adolescents.

Hypothesis 3:

Deviant self-identity will be significantly associated with engagement in subsequent delinquent behavior by adolescents.

Methods

Data

The data for the current study were drawn from the Korean Youth Panel Study (KYPS), which was collected by the Korean National Youth Policy Institute (KNYPI). The KYPS data consisted of six waves of longitudinal data conducted annually from 2003 to 2008. The sampling design employed a stratified multistage cluster, using geographical units as strata and schools and classes to ensure the representativeness of the data (Lee, Kim, and Im 2007). Schools and students were then randomly selected in proportion (Lee, Kim, and Jennings 2020). This sample was representative of the general adolescent population in South Korea, attending school as eighth graders for the first time in the survey in 2003 (Na and Paternoster 2019). Researchers visited the schools and supervised the data collection process to obtain data related to adolescents’ attitudinal, psychological, and behavioral changes while transitioning into early adulthood (Na and Paternoster 2019; Yun, Cui, and Blair 2016). In the initial study, 3,449 students from 104 middle schools were surveyed, and trained interviewers revisited the original respondents until 2008 for a follow-up self-report survey of students and parents (Lee 2018). The total attrition rate from the initial wave to the final wave was 9.8 percent.

The current study excluded cases with missing responses for the indicator (Wave 1), mediators (Wave 2, Wave 3), and outcome measures (Wave 4), which correspond to ages 14 to 17. The total sample used for the analysis was 2,796. To identify systematic bias in the sample selection, an independent samples t-test was conducted using the first-year survey (Lee and Kim 2017). The results revealed statistically insignificant differences, at the 95% confidence level, between the selected sample and samples with missing values in mean differences for engaging in delinquency, informal labeling, deviant self-identity, and control variables.

Measures

Primary delinquency variable (wave 1)

Consistent with previous studies that focused on serious delinquent behaviors rather than status offenses (Bax and Hlasny 2019; Kim and Lee 2019; Kim and Park 2022), seven types of delinquent behaviors were measured and summed to create the measure of primary delinquency using data from Wave 1: severely beating others, taking others’ money or stuff, engaging in gang fights, severely teasing others, stealing others’ money or stuff, threatening, and bullying. Respondents were asked whether they had engaged in these delinquent behaviors in the last year, with ‘no’ coded as 0 and ‘yes’ coded as 1. A score of 0 indicates that respondents have not been involved in any delinquent behavior, while a score of 7 indicates involvement in all types of delinquent behavior (Cronbach’s α = .71).

Mediating variables (wave 2 and wave 3)

Two mediating variables were employed: 1) Informal labeling, and 2) Deviant self-identity. First, informal labeling was measured at Wave 2 using two items assessing respondents’ agreement or disagreement with the statements ‘People around me regard me as a troublemaker’ and ‘People around me regard me as delinquent’ (Kim and Park 2022). Higher scores indicated a greater perception of informal labeling (Cronbach’s α = .89). Second, deviant self-identity was measured at Wave 3 using two questions assessing respondents’ agreement or disagreement with the statements ‘I think of myself as a troublemaker’ and ‘I think of myself as a delinquent’ (Lee 2018; Na and Paternoster 2019). Higher scores reflected a higher level of deviant self-identity (Cronbach’s α = .83).

Subsequent delinquency variable (wave 4)

Subsequent delinquency was measured by asking respondents about their engagement in seven types of delinquent behaviors at Wave 4. These behaviors included severely beating others, taking others’ money or stuff, engaging in gang fights, severely teasing others, stealing others’ money or stuff, threatening, and bullying. A score of 0 indicated non-participation in all delinquent behaviors, while a score of 7 indicated participation in all of these behaviors (Cronbach’s α = .68).

Control variables

To rigorously examine the informal labeling process, indicators that have shown to have a significant effect on adolescents’ delinquent behavior from previous studies were controlled. First, gender was coded as 0 for female adolescents and 1 for male adolescents. Additionally, to account for the association between school achievement and delinquency (Hoffmann, Erickson, and Spence 2013), GPA was measured by using grades in Korean language, English, mathematics, social studies, and science subjects on a 5-point Likert scale (1= very poor level − 5= very good level). The responses were coded so that higher scores represented higher levels of school achievement. Previous studies have consistently demonstrated a significant link between parental attachment and delinquency (Hoeve et al. 2012). Parental attachment was measured using a scale that included six items (e.g., ‘My parents and I spend much time together,’ ‘My parents treat me with love and affection,’… ‘My parents and I understand each other well’). Each item was measured on a 5-Likert scale (1= strongly disagree to 5= strongly agree), and it was coded higher scores to indicate a higher level of parental attachment (α= .81). Similarly, given the well-established link between adolescents’ low self-control levels and delinquent behavior (Pratt and Cullen 2000), low self-control was measured using six items that captured respondents’ impulsivity, tendency to abandon tasks, risk-seeking behavior, and temper (e.g., ‘Even if I have to take an exam tomorrow, I jump into exciting things,’ ‘I abandon a task if it becomes hard and laborious,’ ‘I enjoy risky activities’… “I lose my temper when I get angry”). This variable was coded that a higher score indicating the low self-control they have (α= .72). According to Agnew (1992), strain is the primary source of delinquent behavior. Therefore, strainful experiences were operationalized based on major strains, including the conflict between adolescents and parents, study-related issues, and financial limitations, to control for their effect on delinquent behavior (Jang, Song, and Kim 2014). Four items were used to measure parental strain (e.g., “I get stressed by concerns from my parents about my school grades,” “I get stressed out by disputes with my parents”). Four items were used to measure study strain (e.g., “I get stressed out by poor school grades,” “I get stressed out by assignment or examination preparation”). In addition, three items were used to measure financial strain (e.g., “I get stressed out by the lack of pocket money”). It was coded higher scores to indicate higher levels of strain (α= .68). Furthermore, considering internalizing sources, previous studies have shown a covariation between depression and adolescents’ delinquent behavior (Kofler et al. 2011). Thus, depression was included as a control variable. Depression was measured using six items (e.g., ‘I’m not interested in anything,’ ‘Sometimes I feel very nervous without any reason’… “Sometimes I feel very sad and gloomy without any reason”). It was coded higher scores to indicate a greater number of depressive symptoms (α= .79). Additionally, prior research has highlighted the strong correlation between association with delinquent peers and delinquency (Weerman 2011). Considering the significant association between delinquent peer association and engagement in delinquency (Haynie and Osgood 2005; Thomas 2015), association with delinquent peers was controlled. Delinquent peer association was measured by asking whether close friends were involved in the following delinquent behaviors: drinking, smoking, truancy, severely beating others, taking others’ money or stuff, and stealing others’ money or stuff. Each item was measured as 1 if close friends participated in the corresponding delinquent behavior and 0 if they did not. Since the scores were summed (ranging from 0 to 6), a high score indicated that close friends were involved in a greater number of delinquent behaviors (α= .67). Lastly, previous research suggests that school suspension is associated with an increase in delinquency (Novak 2022). Therefore, to control for its effect on engagement in delinquent behavior, school suspension was measured as whether the respondent had been suspended (0=no, 1=yes).

Analytic strategy

First, univariate analysis was conducted to describe the characteristics of the sample. Also, bivariate correlation analyses were conducted to examine the relationship between variables. Next, to test the hypothesized model in the current study, the structural equation model (SEM) was conducted using Amos 21.0 software (Collier 2020). SEM is a conceptual model used to capture complex and dynamic relationships among observed and unobserved variables (Gunzler et al. 2013). Given the study’s hypotheses, SEM analysis is the most appropriate method for testing mediated effects, as it allows for multiple path analysis (Li 2011). SEM has advantages in controlling for measurement error and capturing mediated effects simultaneously (Preacher, Zyphur, and Zhang 2010). Furthermore, SEM analysis provides more specific model fit information, assessing the consistency of the specified mediational model with the data compared to multiple regression analysis (Bollen and Pearl 2013). To assess the goodness-of-fit of the analysis model, multiple indices were considered, including the χ2 test, Tucker-Lewis Index (TLI), Comparative Fit Index (CFI), and Root Mean Square Error of Approximation (RMSEA) (Bentler 2007). The TLI, also known as the Tucker-Lewis Non-Normed Fit Index (NNFI), is an incremental fit index quantifying the improvement in model fit compared to a null or baseline model. Similarly, the CFI measures the relative improvement in model fit by comparing the target model (the model under evaluation) to a null or baseline model. Both TLI and CFI values close to 1 (e.g., 0.95 or higher) suggest a strong fit between the hypothesized model and the data (Maydeu-Olivares and Garcia-Forero 2010). On the other hand, the RMSEA measures the discrepancy between the observed data and the model-implied covariance structure. It quantifies the average size of the residuals, which are the differences between the observed and predicted covariance, per degree of freedom in the model. The RMSEA typically ranges from 0 to a positive value, with lower values indicating a better fit (Bollen and Pearl 2013). In order to reduce potential bias and enhance the reliability and robustness of the model fit, bootstrapping with 5,000 replicates was conducted. Confirmatory factor analysis (CFA) was employed to evaluate the suitability and effectiveness of the measurement model. In addition, since the variables in the analysis model were operationalized on different scales, standardized coefficients were presented to examine the effects of endogenous indicators in the research model (Liu et al. 2020). Taken together, SEM provides an effective inference framework for the current study by examining the mediation effects of informal labeling and deviant self-identity in the informal labeling process (Gunzler et al. 2013).

Results

Descriptive statistics and measurement model

Table 1 summarizes descriptive statistics of the informal labeling process variables and control variables in the current study. The sample is comprised of both male (51 percent) and female adolescents (49 percent). Correlation analyses between research variables are reported in Table 2. The results indicate that all the key variables including informal labeling variables and control variables were highly correlated in the expected directions. Moreover, the findings, as indicated by the fit indices, demonstrated that the measurement model employed in this study adequately represented the collected data (χ² = 338.245 (p < .001), TLI = .951, CFI = .964, and RMSEA = .033).

Table 1. Descriptive statistics (N = 2,796).

	Mean	SD	Min	Max
Primary delinquency	.601	1.098	0	7
Informal labeling	1.674	.790	1	5
Deviant self-identity	1.842	.722	1	5
Subsequent delinquency	.109	.474	0	6
Gender (0= F, 1= M)	.503	.500	0	1
Parental attachment	3.440	.727	1	5
Low self-control	2.653	.668	1	5
Strainful experiences	2.621	.620	1	5
Depression	2.664	.764	1	5
Delinquent peer association	.934	1.357	0	6
GPA	3.040	.657	1	5
Suspension (0= N, 1= Y)	.010	.098	0	1

Table 2. Bivariate correlations.

	1	2	3	4	5	6	7	8	9	10	11	12
1. Primary delinquency	1	.230**	.192**	.236**	.055**	−.037*	.204**	.047**	.072**	.253**	−.039*	.145**
2. Informal labeling		1	.372**	.188**	.083**	−.165**	.238**	.074**	.112**	.240**	−.141**	.084**
3. Deviant self-identity			1	.149**	.070**	−.200**	.334**	.212**	.241**	.270**	−.200**	.126**
4. Subsequent delinquency				1	.074**	−.054**	.139**	.036	.036*	.254**	−.040*	.204**
5. Gender (0= F, 1= M)					1	−.083**	.049**	−.075**	−.131**	.079**	.068**	.026
6. Parental attachment						1	−.191**	−.176**	−.164**	−.104**	.205**	−.025
7. Low self-control							1	.264**	.235**	.236**	−.205**	.069**
8. Strainful experiences								1	.400**	.081**	−.103**	.032
9. Depression									1	.090**	−.075**	.020
10. DPA										1	−.115**	.150**
11. GPA											1	.002
12. Suspension (0= N, 1= Y)												1

DPA= Delinquency peer association.

P**< .01, P*< .05.

Informal labeling process

Figure 2 presents results from analyzing the structural equation model of the informal labeling process with standardized path estimates. First, it was found that the hypothesized model fits the data well (χ 2 = 212.185 (p < .000), TLI = .974, CFI = .962, and RMSEA = .028).

Figure 2. Standardized path estimates for analysis model.

In terms of Hypothesis 1, primary delinquency was found to be positively related to informal labeling by others (B = .232, p < .001, 95% CI: [.194, 269]). Stated differently, respondents who engaged in primary delinquency were likely to be informally labeled as delinquent by others. Also, supporting Hypothesis 2, informal labeling by others was positively related to the development of deviant self-identity (B = .379, p < .001, 95% CI: [.332, 425]). This indicates that respondents who experienced more informal labeling by others are more likely to develop a higher level of deviant self-identity. Moreover, the result revealed that deviant self-identity was positively related to engaging in subsequent delinquent behavior (B = .054, p < .01, 95% CI: [.038, .072]), which is consistent with Hypothesis 3. Namely, the results indicated that adolescents with a higher level of deviant self-identity are more likely to engage in subsequent delinquent behavior.

The main interest of this study was to examine the validity of the informal labeling process proposed in the labeling theory (Becker 1963; Bernburg 2019; Kubrin, Stucky, and Krohn 2009; Walters 2016). The findings supported the main proposition of the informal labeling process, demonstrating significant paths: primary delinquency → informal labeling by others → development of deviant self-identity → engaging in subsequent delinquent behavior (B= .006, p < .01, 95% CI: [.003, 009]). Furthermore, these pathways of the informal labeling process remained statistically significant even after adjusting for the control variables.

Discussion

Although there have been attempts to examine the association between informal labeling, deviant self-identity, and subsequent delinquent behavior (Bernburg 2019; Kavish, Mullins, and Soto 2016; Lee 2018), research investigating the mediating role of informal labels/altered identity between primary delinquency and secondary delinquency has been limited (Walters 2016; Ward and Tittle 1993). To cover this research gap, using the nationally representative longitudinal data of South Korean adolescents, the primary goal of this study was to examine the validity of the theoretical and empirical scope of the dynamic process of informal labeling, which includes stigmatization by others and the transformation of one’s identity in adolescence.

As expected, the findings presented above generally buttress the informal labeling process proposed by labeling theorists. Specifically, the results indicated the significant mediating role of informal labeling by others and the development of deviant self-identity between primary delinquency and subsequent delinquency. First, it was found that primary delinquency was positively related to informal labeling by others. Second, as hypothesized, the informal labeling by others was in turn positively associated with the development of deviant self-identity. Lastly, it was revealed adolescents’ deviant self-identity was positively related to engaging in subsequent deviant behavior. Additionally, the results indicated the significance of the specified pathway in the informal labeling process. These findings highlight the theoretical importance of informal labeling, which has received less attention compared to formal labeling. Results showed that hypothesized informal labeling process was statistically significant as the expected direction even after controlling other primary indicators related to delinquency. This result is consistent with the informal labeling proposition that informal labeling by others on delinquency increases the alteration of personal identity, which increases the probability of further delinquent behavior (Lemert 1967; Paternoster and Iovanni 1989). In other words, this study supports the core concepts of labeling theory – negative social labels and identity transformation processes – which have received less exploration in previous studies (Bernburg 2019; Walters 2016; Ward and Tittle 1993). Also, the findings suggest that the results from previous studies may be applicable to non-Western samples. While previous studies illustrated adolescents’ reactions to labels might vary in cultural surroundings (Kim and Park 2022; Lee 2018), the findings of the current study support the propositions on informal labeling mechanisms within the South Korean context. Thus, by testing the applicability of the informal labeling model to non-Western samples, the present study contributes to validating the cross-cultural value of the informal labeling model.

Several important implications can be derived from this study. Consistent with the hypotheses of labeling theorists (Lemert 1967; Restivo and Lanier 2015; Schur 1973; Ward and Tittle 1993), informal labeling is associated with unexpected and unintended harmful impact on deviant behavior. Given that deviant self-identity is a salient mediating factor facilitating the effect of informal labeling on subsequent delinquency, interventions to prevent or alleviate the detrimental effect of informal labels might be important (Restivo and Lanier 2015). More specifically, to reduce perceived informal labeling and its associated negative consequences, The Korean Act on the Prevention of and Countermeasures against Violence in Schools was recently revised to address minor delinquency of students informally through school counselors and to prevent the inclusion of delinquent behavior records in academic records (Kim and Park 2022). Similarly, it was found that the school counseling program based on the restorative justice model reduced the negative effects associated with informal labels (Yu, Im, and Jeong 2019). In this regard, Restivo and Lanier (2015) emphasize the importance of adopting a reintegrative approach to protect adolescents from developing a delinquent identity within a school setting. Lastly, previous studies have shown that multisystemic therapy (MST) is effective in preventing subsequent delinquent behavior among juvenile offenders by facilitating changes in their self-identity (Walters 2016). MST demonstrates effectiveness in facilitating the transformation of negative identities to pro-social identities among delinquent adolescents due to its incorporation of significant others involved in the individual’s life (Kavish 2017). Consequently, based on these findings, it is expected that the stigma associated with the development of a deviant-identity by others and engaging in subsequent delinquent behaviors might be reduced though less punitive school intervention policies/programs (Kim and Lee 2019).

Despite the noteworthy findings presented above, this study has several limitations. First, regarding the measurement of informal labeling, the data used in this study did not capture the relative effect of informal labeling from family members, peers, and teachers on the development of deviant-identity. However, Adams et al. (2003) found significant differences in the explanatory power of informal labeling by parents, peers, and teachers in relation to delinquency. Also, they found the effect of informal labeling on delinquency could be varied by the type of delinquency (Lee 2018). Therefore, further research is needed to assess the impact of informal labeling from different sources and explore its influence on various categories of delinquency to gain a more comprehensive understanding of the informal labeling perspective. Lastly, although the research model was designed based on the proposed informal labeling process from labeling theory, it is important to note that the specific temporal direction between the employed variables remains unclear (Walters 2018). This arises from the omission of prior measures of the outcome variables as control variables in the analysis (e.g., informal labeling and deviant self-identity at Wave 1).

In conclusion, this study contributes to filling an empirical gap related to the primary proposition of labeling theory by examining the sequence of informal labeling, which is largely forgotten even with the recent revitalization of labeling theory (Bernburg 2019). Buttressing the ongoing discussion on labeling theory’s advantages in explaining various types of deviance (Na and Paternoster 2019; Restivo and Lanier 2015), the results of the current study provide evidence that supports the proposition of the informal labeling process as an adequate account for the pathway from primary delinquency to engagement in secondary delinquency. This research emphasizes that ‘the informal labeling process deserves further theoretical attention and that future efforts to study the effects of informal labels on delinquency’ (Hayes 1997, 179–180).

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Wonki Lee

Wonki Lee, MA, is a doctoral student in the Department of Sociology and Criminology & Law at the University of Florida. His research interests include criminological theory, juvenile delinquency, and quantitative methods.