Research Article Reviews

344

The Criminal Offending– Self-Esteem Nexus

Which Version of the Self-Esteem Theory Is Supported?

Carrie B. Oser University of Kentucky, Lexington

Traditionalists regard crime to be associated with low self-esteem. A second body of literature suggests a rival hypothesis supporting a relationship between crime and high self-esteem. A multivariate model explored this nexus by exam- ining the impact of demographic characteristics, criminal history, personality traits, and participation in prison programming on self-esteem levels of 134 inmates in two medium-security prisons. Results support the relationship between crime and low self-esteem, with state-raised convicts reporting lower self-esteem levels. In addition, inmates who are emotionally stable, who are less open to novel experiences, and who have participated in counseling or edu- cation during incarceration manifest higher self-esteem.

Keywords: criminal offending–self-esteem nexus; gender; personality traits; prison programming

The criminal offending–self-esteem nexus is one that has been exploredin the criminological literature. The most prominent topic in this line of research supports the relationship between aggressive or violent behavior and low self-esteem (E. Anderson, 1994; Gondolf, 1985; Long, 1990; Oates & Forrest, 1985; Toch, 1969/1993). In contrast, a separate body of litera- ture denotes the controversial view that a narcissistically high level of self- esteem may be related to crime vicariously through the display of aggression (Baumeister, 1997; Baumeister, Boden, & Smart, 1996; Salmivalli, 2001). Self-esteem plays a significant role in relation to the conception of one’s crim- inal identity while incarcerated. Nevertheless, there is a paucity of research on the correlates of inmate self-esteem during incarceration. Living in prison is the ultimate experience of a total institution (Goffman, 1961). Some inmates, labeled “state-raised convicts,” have never successfully navigated

The Prison Journal Volume 86 Number 3

September 2006 344-363 © 2006 Sage Publications

10.1177/0032885506291024 http://tpj.sagepub.com

hosted at http://online.sagepub.com

the free world and are involved in a cycle of repeat offending. There may be a significant relationship between more criminally involved inmates and elevated self-esteem levels. The primary goals of this article are twofold. First, inmate gender differences across criminal history, psychological con- structs, and participation in prison programming will be explored. Second, this article will distinguish which version of the self-esteem theory is sup- ported using a sample of 134 inmates housed in two medium-security pris- ons. Specifically, a multivariate approach is used to regress self-esteem during incarceration on demographic characteristics, criminal history, per- sonality traits, and participation in prison programming.

This article attempts to address several gaps in the literature by explor- ing the correlates of self-esteem using a purposive sample of incarcerated male and female inmates currently serving time in a rural state. First, the previous research on incarcerated individuals is paltry when considering the female sex (Fletcher, Shaver, & Moon, 1993; Johnson, 2002; Pollock, 2002; Sheridan, 1996; Singer, Bussey, Song, & Lunghofer, 1995). This lack of interest in female criminality is apparent despite an increase in the per- centage of women in every phase of the criminal justice system; however, it must be noted that the number of women incarcerated compared to the number of men is relatively small (Pollock, 2002). The disparity between the number of men and women in prison may account for the deficiency of studies on the female incarceration experience. However, female correc- tional facilities do differ from male correctional facilities; therefore, it is relevant to also explore the incarceration experience from a female inmate’s point of view.

In addition to gender differences among inmates, there is a lack of stud- ies exploring the relationship between state-raised convicts and self-esteem. Inmates entering a total institution for the first time may have levels of self- esteem that differ from repeat offenders who have been cycling in and out of the criminal justice system. Thus, the potential covariates of self-esteem, such as criminal history, are examined in the present study.

Researchers know relatively little about the personality correlates of self-esteem (Robins, Tracy, Trzesniewski, Potter, & Gosling, 2001). The self-esteem constructs of inmates have been widely researched, and during the same time frame, personality researchers have conducted a plethora of studies on inmate personality traits. Nevertheless, the conjunction of these two bodies of research is rare in the prison literature.

The prison environment offers few luxuries, and of particular interest is the participation in prison programming for self-improvement. The examination of prison programming in the existing literature focuses primarily on the

Oser / The Criminal Offending–Self-Esteem Nexus 345

impact of prison programs on recidivism rates (Gerber & Fritsch, 1995; Harer, 1995; Inciardi, 1999; Jancic, 1998). Perchance a better measure of the impres- sion made by participation in prison programming would be on the level of self-esteem held by the inmate while incarcerated. In contrast to the conjecture that state-raised convicts have high self-esteem, the prison programming advo- cates argue that low self-esteem is related to criminal offending and that par- ticipation in programming will increase self-esteem to a functional level.

The present study addresses limitations found in previous research by examining the sociodemographic characteristics, criminal history, person- ality profiles, and participation in prison programming among a sample of 134 inmates. Furthermore, the study uses the most current personality scale, the Big Five Factor Model, and the established Rosenberg Self- Esteem Scale. Chi-square and independent samples t tests will be used to examine the differences between male and female incarcerated individuals on a variety of variables. However, it must be stated that because of a small sample size, separate multivariate analyses cannot be run for male inmates and female inmates. An ordinary least squares (OLS) regression model is used to examine the relationship among demographic characteristics, crim- inal history, personality traits, and involvement in prison programming and an inmate’s level of self-esteem. This article is exploratory in nature. If the results indicate a positive relationship between criminal history and self- esteem, the version of the self-esteem theory supporting a link between crime and high self-esteem will be supported. Conversely, if there is a pos- itive relationship between participation in prison programming and self- esteem, the nexus between low self-esteem and crime will be supported.

Review of the Literature

Self-Esteem

As stated above, there is controversial research on the topic of self- esteem. Of importance is the question, what are the repercussions of hav- ing polarized extremes (whether too high or too low) on the self-esteem spectrum? To begin, a lexical definition of self-esteem must be established; therefore, self-esteem refers to one’s self-respect and favorable regard of oneself (Rosenberg, 1979). Results from a meta-analysis suggest that males score higher on global self-esteem than do their female counterparts (Kling, Hyde, Showers, & Buswell, 1999). Previous research suggests that low self-esteem is related to aggressive behavior and violence (E. Anderson,

346 The Prison Journal

1994; Gondolf, 1985; Long, 1990; Oates & Forrest, 1985; Toch, 1969/1993). One version of the low self-esteem theory suggests that low self-esteem is associated with risk taking (Long, 1990). Other research suggests that indi- viduals with low self-esteem may engage in aggressive or violent behavior in an attempt to augment their level of self-esteem (Jankowski, 1991; Toch, 1969/1993). A subculture of violence has also been stated as a result of low self-esteem in that traditionally normative sources of self-esteem are not available, and aggressive behavior is seen as an alternative source.

Gullone, Jones, and Cummins (2000) found that a sample of prison inmates had significantly lower self-esteem levels when compared to a nor- mative student sample. Incarcerated individuals have been shown to have clinically significant problems with self esteem, although no significant dif- ferences were found between males and females (Sheridan, 1996).

However, other empirical findings are suggesting that high self-esteem may be related to involvement in criminal activity. Inflated self-appraisals could lead to violence through the belief that the individual is automatically entitled to the resources of others and could consequently commit criminal acts without remorse. Baumeister and colleagues (1996; Baumeister, 1997), the forerunners of this argument, suggest that there is a discrepancy between internal self-appraisals and external environmental appraisals. The result of this discrepancy is threatened egotism, which subsequently creates a choice point of rejection or acceptance of the external appraisal. When the appraisal is accepted, negative emotions toward oneself occur, and with- drawal is the final result. On the contrary, one can reject the appraisal and maintain one’s self-appraisal, which results in negative emotions toward the source of the threat. The outcome of the appraisal route is aggression or violence (Baumeister et al., 1996).

From this vantage point, it is believed that individuals who aggress can be characterized as displaying a belief in their superiority, not inferiority (Baumeister et al., 1996). A concept known as defensive self-esteem under- lies this notion of superiority and has been addressed in the self-esteem lit- erature to distinguish between those with genuine self-esteem and those with problematically high self-esteem. This concept suggests that some individu- als may present an inflated concept of the self to avoid the conscious or unconscious realization that they have a low self-esteem (Harder, 1984; Salmivalli, 2001). In this respect, self-esteem may reflect the concept of self- presentation, or the manner in which the individual would like to present himself or herself to others (Baumeister, Tice, & Hutton, 1989; Salmivalli, 2001). Another possibility is presented by the psychodynamic view in which narcissistically disturbed individuals deny negative characteristics to maintain

Oser / The Criminal Offending–Self-Esteem Nexus 347

a positive self-view (Salmivalli, 2001). An insecure arrogance may cause aggression, which results in criminal activity (Baumeister, 1997).

The majority of studies assessing self-esteem do not use inmates as their unit of analysis. This population is distinctively different from other popu- lations because of the imposed limitations on their freedom. In particular, no studies addressing Baumeister’s (1997, 2001; Baumeister et al., 1989, 1996) notion of threatened egotism were found in the criminal justice liter- ature. The contradictory evidence on the self-esteem–crime nexus and the lack of an appropriate inmate venue demonstrates the need to further exam- ine self-esteem among incarcerated populations. Therefore, an overview of the literature on the relationship among self-esteem and criminal history, personality traits, and prison programming is discussed below.

Criminal History

The ecology of the prison environment is not conceptually distinct from the outside world because both communities are composed of hierarchies. However, because of the nature of a total institution, the procedures used by individuals to maneuver in the prison world are quite different from behav- iors used outside the prison walls. Criminal history plays a critical role in the prison assimilation process. In regard to the hierarchy of inmates housed in a prison setting, the state-raised convict is of particular interest. State-raised convicts are individuals at the top of the inmate hierarchy who take advantage of the general citizenry of the prison population (Johnson, 2002). Convicts are characterized by their use of violence, lack of empathy for others, and tough demeanor, which in prison translates into power and a high status (Johnson, 2002). State-raised convicts have woven in and out of the criminal justice system throughout their lives and seem to be just as responsive to life in prison as to life in the free community.

Most likely, these predatory types of individuals lack conventional bonds to society and were raised in economically disadvantaged areas. Impoverished communities are often notorious for violence associated with gang mem- bership and/or the related drug trade. It could be argued that the state-raised convict, who normally serves short sentences for drug-related offenses, has brought this violent predisposition from the streets to the prison commu- nity. Therefore, the overall demeanor of the modern prison has become increasingly violent because of the value system associated with state- raised convicts.

Although the state-raised convict is destined to fail in the free world, he or she is unafraid of the prison environment. In relation to self-esteem, the

348 The Prison Journal

prison experience may have provided the persistent felon with the forum to demonstrate how tough he or she is, thereby creating an elevated self- esteem. In fact, studies with prisoners demonstrate that the self-esteem of the inmates increased with longer time served (Greve & Enzmann, 2003; Wormith, 1984). As an example of defensive self-esteem, the state-raised convict may endorse his or her criminal accomplishments and embrace the feat of survival in prison or in other total institutions such as juvenile facil- ities (Johnson, 2002). Moreover, the serious criminal may have an inflated self-image because it reduces the pain of being repeatedly rejected by society (Baumeister, 2001; Salmivalli, 2001; Walters, 1990). According to the theory posed by Baumeister and colleagues (1996; Baumeister, 1997), state-raised convicts would be well-equipped to deal with threatened ego- tism in the violent prison environment. When a combination of a highly favorable self-image and ego threat occurs, the inmate may react aggres- sively toward the source of the threat. Based on this theoretical argument, it is hypothesized that the extent to which an inmate is a state-raised con- vict (operationalized as possessing a more extensive criminal history) will be positively associated with his or her level of self-esteem. Specifically, inmates convicted of violent or drug offenses are expected to have signifi- cantly higher levels of self-esteem. In addition, prisoners with long sen- tence lengths, more previous juvenile convictions, and a greater number of prior incarcerations are expected to manifest higher self-esteem levels while incarcerated.

This hypothesis tests the high self-esteem version of the theory, which emphasizes the relationship between criminal offending and elevated self- esteem levels among an incarcerated population. Hence, if the criminal history of a state-raised convict is influencing a higher degree of self- esteem, then the next important step is to move beyond the environmental context to explore the role of individual personality traits. Thus, the next section outlines some of the key findings in the personality literature.

Personality Traits

Personality traits are logically important in connection to the other individual-level differences research on self-esteem. The notion of a hierarchal trait configuration dates back to Allport’s (1937) citation of a three-tiered trait hierarchy (Guastello, 1993). One of the shortcomings of the assess- ment of personality traits is seen in the lack of updated personality survey instruments. Therefore, outdated personality taxonomies are replaced by the Big Five Inventory-44 (BFI-44), created by Oliver John (1995). The

Oser / The Criminal Offending–Self-Esteem Nexus 349

BFI-44 is derived from Goldberg’s Big Five Factor Inventory and is designed to measure the five orthogonal factors of extraversion, agreeable- ness, conscientiousness, emotional stability (neuroticism), and openness (John, 1995). The BFI-44 is advantageous because it is a validated, com- prehensive model of personality that subsumes multiple personality traits.

Using this instrument to distinguish criminal personality traits would contribute to the understanding of self-esteem. The majority of research in this area uses college student samples and only focuses on the Big Five per- sonality correlates of self-esteem (Keller, 1999; Kwan, Bond, & Singelis, 1997; Robins et al., 2001). Most research reveals that self-esteem has a strong, positive correlation with emotional stability (neuroticism), a mod- erately positive correlation with extraversion and conscientiousness, and a weak, positive correlation with agreeableness and openness (Jackson & Gerard, 1996; Pullman & Allik, 2000; Robins et al., 2001).

According to research by Robins and colleagues (2001), approximately 34% of the variance in self-esteem is explained by the Big Five. Internet data collection revealed the finding that individuals with high self-esteem tend to score high on the Big Five dimensions of extraversion, agreeable- ness, conscientiousness, emotional stability (neuroticism), and openness (Robins et al., 2001). There are no known studies exploring this issue using an incarcerated population. This existing literature of nonincarcerated indi- viduals would suggest that inmates scoring higher on the personality dimensions of extraversion, agreeableness, conscientiousness, emotional stability, and openness to experience will manifest higher levels of self- esteem while incarcerated.

Prison Programming

An interesting feature of correctional settings is the role of prison pro- gramming, although this curriculum has drastically diminished in the past decade. Prison programming, as a utilitarian rationale for punishment, was popular in the United States in the 1960s and 1970s, until America began to endorse the war on drugs through an increase in get-tough policies. Despite the lack of educational and vocational training among incarcerated programs, inmate involvement in prison programming is low. A study by Morash, Haarr, and Rucker (1994) indicates that 20.6% of women and 19.9% of men report receiving any vocational training while incarcerated. Moreover, existing vocational training is often gender stereotyped, and it does not sufficiently prepare individuals for employment (King, 1993; Morash et al., 1994).

350 The Prison Journal

The relationship between prison program participation and self-esteem is usually discussed within the context of studying the recidivism-reducing role of prison programs (Inciardi, 1999). This measure of success can often be problematic for those endorsing a harm-reductionist perspective. In addition, some scholars argue that prison programming has both promise and value (D. B. Anderson, 1981; D. B. Anderson, Schumacker, & Anderson, 1991; Schumacker, Anderson, & Anderson, 1990), whereas others think that it does not work (Martinson, 1974; Sullivan, 1990). Perhaps a better alternative is to look at the role of prison programs in increasing one’s pos- itive self-image. Participation in behavioral or cognitive programming can create a sense of pride among inmates and make them feel valued (Pollock, 2002). To expect an appreciable reduction on recidivism rates is unrealistic because the programs may actually attempt too much (Johnson, 2002). Wooldredge (1999) found that psychological well-being is positively influ- enced by participation in institutional prison programs.

This vantage point emphasizes the negative relationship between crimi- nal offending and self-esteem. Essentially, the argument supports the crim- inal offending–low self-esteem notion because it is possible that inmates, who may be deemed as prey on the prison yard, may participate in pro- gramming to avoid those dangerous situations. A study by Owen (1998) on female inmates revealed that although the gang or prison clique may be the state-raised convict’s program, inmates may prefer constructive educational or behavioral programs that offer both routine activities and protection from the predatory inmates on the yard. As alluded to above, there is a large body of research that supports the relationship between criminal offending and low levels of self-esteem; therefore, it may be posited that individuals who fall into the unhealthily low end of the self-esteem spectrum may benefit more from educational and vocational programs than state-raised convicts. Based on the extant prison programming literature, it is expected that par- ticipation in prison programming is positively associated with self-esteem. Specifically, the greater the participation in prison programming, the higher the inmates’ self-esteem while incarcerated.

This hypothesis tests the low self-esteem version of the theory that empha- sizes the relationship between engagement in criminal activities and low self- esteem. The utopian goals of rehabilitation programs can never be achieved; nonetheless, rehabilitative goals should be revisited to create more positive inmate characteristics. The goals should be redefined in terms of committing to the program, following through, alleviating boredom, creating a caring envi- ronment, developing a sense of self-efficacy, and increasing self-esteem.

Oser / The Criminal Offending–Self-Esteem Nexus 351

Method

Sample

Between 1997 and 1998, 146 inmates at two medium-security prisons completed a self-administered survey. The survey was designed to assess various social–psychological concepts and other demographic characteris- tics during incarceration. Inmates for this purposive sample were recruited through recruitment flyers, requesting participation, sent to the prisons. Those ineligible for the sampling frame were inmates in disciplinary and maximum-security units. Participation was voluntary, and no incentives were offered. The research procedures were approved after a full review by the appropriate ethical institutional review boards within a university setting.

Measurement of Dependent Variable—Self-Esteem

The dependent variable is the social–psychological characteristic of self- esteem. The Rosenberg Self-Esteem Scale, a 10-item scale, is used to mea- sure individual inmates’ perceived self-worth. This unidimensional measure of self-esteem has previously reported internal consistency and validity (Rosenberg, 1965). The interitem reliability for the 10 measures of self- esteem on these data is also strong (α = .84). The self-esteem scale is an additive scale ranging from a minimum of 10 (low self-esteem) to a maxi- mum of 40 (high self-esteem).

Demographic Measures

The influence of several demographic measures on inmate self-esteem is noted in the literature, and they are accordingly included in the multivari- ate model as control variables. First, gender is included, with males coded 0 and females coded 1. The age of the respondent is measured in number of years. Because of the small number of Asians or Pacific Islanders and Latinos, a dichotomous variable is created for race (0 = White, 1 = non- White). Marital status and education are also included as control variables. Divorced, widowed, or single inmates are coded 0, whereas legally married inmates are coded 1. Educational attainment is used as a general proxy for socioeconomic status (0 = did not finish high school, 1 = completed high school or GED).

352 The Prison Journal

Criminal History

The state-raised convict component of the model, as measured by crim- inal history, includes the type of offense of the current conviction, the length of sentence, the number of previous juvenile convictions, and the number of the previous adult convictions. Offense type is measured through the creation of a set of dichotomous variables indicating whether the current offense was a violent crime (0 = no, 1 = yes), a sex crime (0 = no, 1 = yes), a property crime (0 = no, 1 = yes), or a drug crime (0 = no, 1 = yes). Inmates may fall into more than one offense category because they may be serving time for multiple offenses.

Length of sentence is recorded as the natural log of the number of months. Inmates serving life sentences are capped at the longest sentence length (1,000 months). A log transformation of this indicator is appropriate because the distribution of sentence length is positively skewed. Juvenile criminal history is the self-reported number of juvenile convictions. Finally, the notion of a state-raised convict is also measured by the number of pre- vious adult convictions, which serves as a proxy for repeat offenders.

Personality Traits

Personality traits, the third category of interest, are measured by the BFI-44 (John, 1995). This newly condensed version of 44 question items from Goldberg’s original Big Five Factor Inventory, a well-known self-report mea- sure of the Big Five, is used to measure the nomothetic dimensions of extra- version, agreeableness, conscientiousness, emotional stability (neuroticism), and openness among incarcerated populations (Costa & McCrae, 1985). After conducting reliability analyses on items representing the BFI-44, the follow- ing alpha coefficients for the male inmates are revealed for each of the five personality traits: extraversion (α = .6839), agreeableness (α = .6884), con- scientiousness (α = .7270), emotional stability (neuroticism; α = .7957), and openness (α = .6861). In contrast, the following alpha coefficients for the female inmates for each of the five personality traits are as follows: extraver- sion (α = .7295), agreeableness (α = .7149), conscientiousness (α = .6866), emotional stability (neuroticism; α = .8254), and openness (α = .7066).

Prison Programming

Prison programming is the fourth component of interest in explaining the level of self-esteem among incarcerated individuals. To examine the

Oser / The Criminal Offending–Self-Esteem Nexus 353

impact of prison program participation, a series of five dichotomous vari- ables (where 0 = no, 1 = yes) was created. The programs of interest include psychological counseling, violence prevention programs, vocational train- ing, involvement in 12-step programs such as Alcoholics Anonymous or Narcotics Anonymous, and the completion of high school or a GED during incarceration.

Analytic Strategy

Chi-square and independent samples t tests are used to identify signifi- cant differences between male (n = 76) and female inmates (n = 58) across the categorical and continuous study variables. Although the sample size for this research is too small to run separate gender multivariate models, the bivariate analyses provide rationale for future research. Results from the correlation matrix (not shown) demonstrated that none of the independent variables correlated so highly as to imply problems with multicollinearity. An OLS regression model (n = 134) is used to detect the relationship among demographic characteristics, state-raised inmate measures, person- ality traits, and participation in prison programming on the level of self- esteem while incarcerated. Twelve cases were excluded from the analyses because of missing data.

Results

Bivariate Analysis

Table 1 displays the results of the chi-square and independent samples t tests. The majority of inmates in the sample are male (57%), but females are well represented for a study on incarcerated individuals (43%). The inmates range in age from 20 to 61 years (M = 35.19, SD = 8.88). Consistent with the racial breakdown of incarcerated populations in this geographic location, 35% of the participants are non-White. Approximately 25% of inmates report being married, and this is the only demographic characteristic that significantly differs between male and female inmates. Females are significantly more likely to be legally married (33% of females vs. 18% of males). The inmate population is atypically overly educated, with the vast majority reporting that they have completed high school or a GED (M = 0.81, SD = 0.40). More than 44% of inmates report having com- pleted some college. The well-educated sample of prison inmates in this

354 The Prison Journal

Oser / The Criminal Offending–Self-Esteem Nexus 355

study may be the result of a self-selection bias to participate in voluntary research.

Several variables are used to measure the concept of the state-raised inmate. The majority of inmates are convicted of a violent crime (49%), followed by property crimes (32%) and/or drug offenses (25%). Inmates serving time for multiple crimes are accordingly included in multiple offense categories.

Table 1 Descriptive Statistics and Results of Bivariate Analysis

Male Female Total Inmatesa Inmatesb Samplec

Self-Esteem Scale (range = 10.00-40.00)d 18.83 19.29 19.03 Demographics

Aged 34.35 36.30 35.19 Non-White (1 = yes) 0.35 0.35 0.35 Married (1 = yes) 0.18 0.33 0.25* High school diploma or GED (1 = yes) 0.78 0.84 0.81

Criminal history (state-raised inmate) Violent offense (1 = yes) 0.59 0.36 0.49** Sex crime offense (1 = yes) 0.05 0.02 0.04 Drug offense (1 = yes) 0.18 0.33 0.25* Property crime (1 = yes) 0.32 0.33 0.32 Length of sentence (in months)d 174.16 113.30 145.47* Number of juvenile incarcerationsd 2.26 0.56 1.52** Number of adult incarcerationsd 3.78 3.79 3.79

Personality traits Extraversion (range = 11.00-31.00)d 21.63 22.30 21.92 Agreeableness (range = 19.00-36.00)d 27.54 29.06 28.20 Conscientiousness (range = 18.00-72.00)d 27.66 29.24 28.34 Emotional stability (neuroticism; 19.63 20.26 19.90

range = 10.00-32.00) Openness (range = 18.00-36.00)d 28.80 27.96 28.44

Prison programming Psychological counseling (1 = yes) 0.17 0.41 0.28** Violence prevention (1 = yes) 0.13 0.16 0.14 Vocational training (1 = yes) 0.26 0.31 0.28 12-step program (1 = yes) 0.39 0.57 0.47* High school or GED (1 = yes) 0.41 0.22 0.33*

a. n = 76. b. n = 58. c. n = 134. d. Independent samples t tests were used for continuous variables. *p < .05. **p < .01.

Independent samples t tests indicate that males are overrepresented among inmates serving time for a violent offense, whereas females are overrepre- sented among inmates serving time for a drug conviction. Inmates are serv- ing relatively long sentences (M = 145.47, SD = 3.29). In addition, the sample reported an average of 1.52 juvenile convictions and an average of 3.79 adult convictions. Gender differences are found with males possessing more of the characteristics of a state-raised inmate. Specifically, male inmates report significantly longer sentences and a greater number of pre- vious juvenile convictions than do female inmates.

Participants tend to be high, above the scale midpoint, on all personality traits, with the exception of neuroticism. None of the Big Five personality trait variables differs significantly across gender lines.

The most popular programs among inmates are 12-step programs (47%) and educational programs (33%). Significant differences are found between men and women for two of the five prison programming variables. Females are overrepresented in their participation in psychological counseling but are underrepresented in their involvement in educational programs.

Multivariate Analysis

The OLS model estimating the effects of demographic characteristics, state-raised convict measures, personality traits, and prison programming on inmate level of self-esteem during incarceration is displayed in Table 2. Education is the only significant demographic characteristic related to self- esteem. Specifically, individuals who have completed high school or a GED have significantly higher levels of self-esteem while incarcerated.

In the items measuring the concept of a state-raised convict, three of the variables have a significant relationship with self-esteem. The positive rela- tionship between the criminal history measures and self-esteem provide some support for the low self-esteem and crime nexus. Specifically, inmates convicted of violent or drug offenses had significantly lower levels of self- esteem than inmates convicted of other offenses. Also, contrary to the hypothesis linking criminal history to high self-esteem, a negative relation- ship was found between the number of previous adult convictions and self- esteem. Specifically, an increase in the number of adult convictions is associated with a decrease in the inmate’s level of self-esteem. Length of sentence and the number of juvenile incarcerations did not have significant relationships with inmate self-esteem.

Two of the Big Five personality dimensions are important predictors of self-esteem. First, emotional stability has a direct, positive effect on the

356 The Prison Journal

Oser / The Criminal Offending–Self-Esteem Nexus 357

inmate’s level of self-esteem. To state it another way, neurotic incarcerated individuals have lower levels of self-esteem. Second, as the Big Five per- sonality trait of openness decreases, self-esteem significantly increases among incarcerated individuals. Essentially, the less open the inmate is

Table 2 Ordinary Least Squares Model Estimating the Correlates

of Demographic Characteristics, Criminal History, Personality Traits, and Prison Programming on Inmate Level

of Self-Esteem During Incarceration

b SE

Demographics Gender –0.226 0.868 Age 0.018 0.048 Non-White –0.020 0.817 Married 0.095 0.948 High school diploma or GED 2.001* 1.025

Criminal history (state-raised inmate) Violent offense –3.270** 1.080 Sex crime offense –3.306 2.280 Drug offense –3.809** 1.106 Property crime –0.459 0.970 Length of sentence 0.061 0.388 Number of juvenile incarcerations 0.094 0.099 Number of adult incarcerations –0.090* 0.043

Personality traits Extraversion –0.119 0.101 Agreeableness –0.124 0.088 Conscientiousness 0.096 0.080 Emotional stability (neuroticism) 0.413** 0.077 Openness –0.214* 0.096

Prison programming Psychological counseling 1.795* 0.910 Violence prevention 1.153 1.114 Vocational training –0.700 0.867 12-step program –0.255 0.779 High school or GED 1.823* 0.912

Constant 19.904 F score 4.332** Adjusted R 2 0.355

Note: n = 134. *p < .05, one-tailed. **p < .01, one-tailed.

with other inmates and correctional staff, the higher the inmate’s level of self-esteem.

The last component of interest is prison programming, of which two of the five explored programs have significant relationships with self-esteem. Both psychological counseling and participation in educational programs have a significant, positive effect on the inmate’s level of self-esteem while incarcerated. As the percentage of inmates participating in psychological counseling or educational programs increases, so too does the level of self- esteem while incarcerated. The positive relationship between involvement in psychosocial or educational programs during incarceration and self- esteem substantiate the low self-esteem version of the theory.

The fully specified model is statistically significant (F = 4.332) at the p < .01 level. The model accounts for approximately 36% of the variance in the level of self-esteem of incarcerated inmates.

Discussion

There are two viewpoints of the nexus between crime and self-esteem. The traditional view purports that low self-esteem is a correlate of violence and aggressive behavior (E. Anderson, 1994; Gondolf, 1985; Long, 1990; Oates & Forrest, 1985; Toch, 1969/1993), whereas the controversial viewpoint sug- gests that individuals with highly favorable self-appraisals are the ones most likely to lead to criminal offending (Baumeister, 1997; Baumeister et al., 1996; Salmivalli, 2001). The present study explores the criminal offending– self-esteem nexus by adding to the current body of self-esteem literature in three ways. First, an underpursued area of self-esteem is investigated— male and female self-esteem during incarceration. Specifically, individuals embedded in total institutions are specifically vulnerable to the ecological dimensions of prison survival. Understandably, a person’s self-esteem while incarcerated should differ from that in the real world because various factors differ in these two environments. The notion of a state-raised convict in relation to high self-esteem has not been studied. A state-raised convict, or someone who has been in the justice system repeatedly since youth, may have a defensive self-esteem or may be narcissistically disturbed. This study acknowledges that criminal history may be related to the elevated self-esteem of incarcerated individuals. Specifically, it explores if a state-raised convict, operationalized as engagement in a high level of criminal offending, is related to higher levels of self-esteem. Moreover, male correctional facilities are considerably different from female correctional facilities (Johnson, 2002;

358 The Prison Journal

Pollock, 2002). Therefore, the present study explores self-esteem issues of those individuals in the controlled environment of the prison and conducts bivariate analyses to explore the incarceration experience across gender lines.

Second, personality traits and self-esteem levels are two important lines of research on individual differences that have rarely been connected (Robins et al., 2001). Although several studies explore the correlation between personality and self-esteem (Jackson & Gerard, 1996; Keller, 1999; Kwan et al., 1997; Pullman & Allik, 2000), only one study examines the predictive power of the Big Five dimensions in predicting self-esteem (Robins et al., 2001). The present line of research is innovative in that it acknowledges the personality correlates of self-esteem among an incarcer- ated population using the BFI-44.

Third, the efficacy of prison programming in reducing the recidivism rates is debated (Inciardi, Martin, Butzin, Hooper, & Harrison, 1997; Martinson, 1974; Wilson, 1980), albeit it is important to recognize that par- ticipation in prison programming may have other desirable outcomes, such as increased self-esteem. If there is a link between criminal offending and low self-esteem, perhaps the objective of prison programming should be an increase in self-esteem while incarcerated. The relationship between par- ticipation in prison programming and self-esteem is explored in the current research to examine the low self-esteem–criminal offending version.

Overall, the low self-esteem–criminal offending relationship is sup- ported using a sample of 134 incarcerated individuals. This finding sub- stantiates previous research by suggesting that the engagement in criminal activities is associated with low levels of self-esteem (E. Anderson, 1994; Gondolf, 1985; Long, 1990; Oates & Forrest, 1985; Toch, 1969/1993). State-raised convicts are not displaying defensive self-esteem and may not be narcissistically disturbed while incarcerated. On the contrary, inmates serving time for violent criminal acts or for drug offenses have significantly lower self-esteem. The state-raised convict may be able to successfully nav- igate the prison underworld (Johnson, 2002), but he or she maintains a low self-esteem while doing so. The perceived necessity to use force to produce desirable outcomes, in the case of violent offenders, or the ability to reduce or abstain from substance use, in the case of the drug users, may lead the inmate to experience a feeling of worthlessness during incarceration. In addition, state-raised convicts, or those with longer sentence lengths and greater numbers of juvenile convictions, are significantly more likely to have lower levels of self-esteem. Perhaps the inability of these career crim- inals to successfully navigate the free world manifests into lower levels of self-esteem.

Oser / The Criminal Offending–Self-Esteem Nexus 359

There was partial support for the hypothesis suggesting that inmates who score higher on the personality dimensions of extraversion, agreeable- ness, conscientiousness, emotional stability (neuroticism), and openness to experience will have higher levels of self-esteem. To be precise, only two of the Big Five personality traits were significantly related to self-esteem. Consistent with previous research on nonincarcerated populations (Jackson & Gerard, 1996; Pullman & Allik, 2000; Robins et al., 2001), inmates with greater emotional stability have higher levels of self-esteem while incarcer- ated. Emotionally stable inmates may be psychologically better equipped to handle the incarceration experience. Conversely, the studies on nonincar- cerated populations show a positive association between openness and self- esteem, whereas the present study found a negative relationship (Jackson & Gerard, 1996; Pullman & Allik, 2000; Robins et al., 2001). Because of the nature of the incarceration experience, it can be argued that these personal- ity traits aid the inmate in adapting to the environment of a total institution; thereby, inmates less open to novel experiences may be rule abiders. In this situation, it is logical that inmates who are less open to novel experiences during incarceration may manifest higher self-esteem levels.

The hypothesis suggesting a positive relationship between participation in prison programming and self-esteem was also partially supported. Again, this provides support for the link between crime and low self-esteem. Specifically, two of the five programs were significant predictors of self- esteem. The more inmates who participate in psychological counseling and educational programs, the higher the self-esteem levels during incarcera- tion. It must be noted that individuals participating in programming may be conceptually different than those not participating in terms of self-esteem lev- els. It is plausible that low self-esteem inmates are either targeted within the prison environment for behavioral or educational programs or are more likely to participate in prison programming.

Although these findings are valuable, there are several limitations that must be addressed. Of primary importance is sample size. OLS regression is sensitive to sample size, and an increase in the number of cases could substantially increase the explanatory power of the multivariate model. In addition, external validity is limited because the sample was not randomly selected. Data were collected from prisoners who volunteered for the study in purposively selected prisons in one state.

Cross-sectional data were used in the present study. Although the purpose of this study was to explore the correlates of self-esteem during incarceration, causal inferences cannot be drawn regarding the extent to which self-esteem is a consequence of criminal history, personality traits, and participation in

360 The Prison Journal

Oser / The Criminal Offending–Self-Esteem Nexus 361

prison programming. One can only deduce from this investigation correla- tion, not causation.

Future research could improve on similar studies through the imple- mentation of several changes. First, the paucity of research concerning the relationship between criminal offending and self-esteem while incarcerated could better operationalize the notion of a state-raised convict and partici- pation in prison programming. Additional measures of an inmate’s family history and the engagement and duration of involvement in programs could be important correlates of self-esteem during incarceration. The addition of more precise measures of the concepts, in addition to an increase in sample size, could be fruitful for future studies.

Overall, this study provides an important reexamination of the criminal offending–self-esteem nexus. Although no definitive answer concerning the correlates of self-esteem while incarcerated were reached, the findings do contribute to the current body of literature of self-esteem. Specifically, the link between crime and low self-esteem is further supported. Findings sug- gest that variables other than demographic characteristics are most apt at explaining the variance in the degrees of inmate self-esteem while incarcer- ated. Inmates with extensive criminal histories were found to have signifi- cantly lower self-esteem levels. In addition, there was a significant correlation between certain personality traits and self-esteem. Inmates possessing emo- tional stability and those who were less open to novel experiences exhibited lower levels of self-esteem. Finally, taking part in behavioral or educational programs during incarceration is associated with elevated self-esteem levels. In the end, this study’s findings serve as a barometer in assessing the level of self-esteem during incarceration and how criminal history, personality traits, and prison programming are associated with self-esteem.

References

Allport, G. W. (1937). Personality: A psychological interpretation. New York: Henry Holt. Anderson, D. B. (1981). The relationship between correction education and parole success.

Journal of Offender Counseling, Services, and Rehabilitation, 5, 13-25. Anderson, D. B., Schumacker, R. E., & Anderson, S. L. (1991). Release characteristics and

parole success. Journal of Offender Rehabilitation, 17, 133-145. Anderson, E. (1994). Code of the street. New York: Norton. Baumeister, R. F. (1997). Evil. Inside human cruelty and violence. New York: Freeman. Baumeister, R. F. (2001). Violent pride. Scientific American 284(4), 96-101. Baumeister, R. F., Boden, J. M., & Smart, L. (1996). Relation of threatened egotism to vio-

lence and aggression: The dark side of self-esteem. Psychological Review, 103, 5-33.

362 The Prison Journal

Baumeister, R. F., Tice, D., & Hutton, D. (1989). Self-presentational motives and personality differences in self-esteem. Journal of Personality, 57, 547-579.

Costa, P. T., Jr., & McCrae, R. R. (1985). The NEO personality inventory manual. Odessa, FL: Psychological Assessment Resources.

Fletcher, B. F., Shaver, L. D., & Moon, D. G. (1993). Women prisoners: A forgotten popula- tion. Westport, CT: Praeger.

Gerber, J., & Fritsch, E. J. (1995). Adult academic and vocational correctional education programs: A review of recent research. Journal of Offender Rehabilitation, 22, 199-242.

Goffman, E. (1961). Asylums: Essays on the social situation of mental patients and other inmates. New York: Doubleday.

Gondolf, E. W. (1985). Men who batter. Holmes Beach, FL: Learning Publications. Greve, W., & Enzmann, D. (2003). Self-esteem maintenance among incarcerated young males:

Stabilisation through accommodative processes. International Journal of Behavioral Development, 27(1), 12-20.

Guastello, S. J. (1993). A two-(and-a-half)-tiered trait taxonomy. American Psychologist, 48, 1298-1299.

Gullone, E., Jones, T., & Cummins, R. (2000). Coping styles and prison experience as predic- tors of psychological well-being in male prisoners. Psychiatry, Psychology, and Law, 7(1), 170-181.

Harder, D. (1984). Character style of the defensively high self-esteem man. Journal of Clinical Psychology, 40, 26-35.

Harer, M. D. (1995). Recidivism among federal prisoners released in 1987. Journal of Correctional Education, 46(3), 98-128.

Inciardi, J. A. (1999). Criminal justice (6th ed.). Fort Worth, TX: Harcourt Brace. Inciardi, J. A., Martin, S. S., Butzin, C. F., Hooper, R. M., & Harrison, L. D. (1997). An effec-

tive model of prison-based treatment for drug-involved offenders. Journal of Drug Issues, 27, 261-278.

Jackson, L. A., & Gerard, D. A. (1996). Diurnal types, the “Big Five” personality factors, and other personal characteristics. Journal of Social Behavior and Personality, 11, 273-283.

Jancic, M. (1998). Does correctional education have an effect on recidivism? Journal of Correctional Education, 49(4), 152-161.

Jankowski, M. S. (1991). Islands in the street: Gangs and American urban society. Berkeley: University of California Press.

John, O. P. (1995). Big five inventory. Unpublished report, Institute of Personality and Social Research.

Johnson, R. (2002). Hard time: Understanding and reforming the prison (3rd ed.). Belmont, CA: Wadsworth.

Keller, T. (1999). Images of the familiar: Individual differences and implicit leadership theo- ries. Leadership Quarterly, 10, 589-607.

King, E. O. (1993). African American males in prison: Are they doing time or is the time doing them? Journal of Sociology and Social Welfare, 20(4), 9-27.

Kling, K. C., Hyde, J. S., Showers, C. J., & Buswell, B. N. (1999). Gender differences in self- esteem: A meta-analysis. Psychological Bulletin, 125, 470-500.

Kwan, V. S. Y., Bond, M. H., & Singelis, T. M. (1997). Pancultural explanations for life satis- faction: Adding relationship harmony to self-esteem. Journal of Personality and Social Psychology, 73, 1038-1051.

Long, D. E. (1990). The anatomy of terrorism. New York: Free Press.

Martinson, R. (1974). What works? Questions and answers about prison reform. Public Interest, 35, 22-54.

Morash, M., Haarr, R. N., & Rucker, L. (1994). A comparison of programming for women and men in U.S. prisons in the 1980s. Crime & Delinquency, 49, 197-221.

Oates, R. K., & Forrest, D. (1985). Self esteem and early background of abusive mothers. Child Abuse and Neglect, 9, 89-93.

Owen, B. (1998). “In the mix”: Struggle and survival in a women’s prison. Albany: State University of New York Press.

Pollock, J. M. (2002). Women, prison, and crime. Belmont, CA: Wadsworth. Pullman, H., & Allik, J. (2000). The Rosenberg self-esteem scale: Its dimensionality, stability, and

personality correlates in Estonian. Personality and Individual Differences, 28, 701-715. Robins, R. W., Tracy, J. L., Trzesniewski, K., Potter, J., & Gosling, S. D. (2001). Personality

correlates of self-esteem. Journal of Research in Personality, 35, 463-482. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton

University Press. Rosenberg, M. (1979). Conceiving the self. New York: Basic Books. Salmivalli, C. (2001). Feeling good about oneself, being bad to others? Remarks on self-

esteem, hostility, and aggressive behavior. Aggression and Violent Behavior, 6, 375-393. Schumacker, R., Anderson, D. B., & Anderson, S. L. (1990). Vocational and academic indica-

tors of parole success. Journal of Correction Education, 41, 8-12. Sheridan, M. J. (1996). Comparison of the life experiences and personal functioning of men

and women in prison. Families in Society: The Journal of Contemporary Human Services, 6, 123-134.

Singer, M. E., Bussey, J., Song, L., & Lunghofer, L. (1995). The psycho-social issues of women serving time in jail. Social Work, 40, 103-113.

Sullivan, L. E. (1990). The prison reform movement: Forlorn hope. Boston: Twayne. Toch, H. (1993). Violent men: An inquiry into the psychology of violence. Washington, DC:

American Psychological Association. (Original work published 1969) Walters, G. D. (1990). The criminal lifestyle: Patterns of serious criminal conduct. Newbury

Park, CA: Sage. Wilson, J. Q. (1980). “What works?” revisited: New findings on criminal rehabilitation. Public

Interest, 61, 3-17. Wooldredge, J. D. (1999). Inmate experiences and psychological well-being. Criminal Justice

and Behavior, 26, 235-250. Wormith, J. S. (1984). Attitude and behavior change of correctional clientele. Criminology,

22, 595-618.

Carrie B. Oser is an assistant professor in the Sociology Department and a faculty associate of the Center on Drug and Alcohol Research at the University of Kentucky. Her research inter- ests include HIV risk behaviors among individuals involved in the criminal justice system, health services research, and rural substance use and abuse.

Oser / The Criminal Offending–Self-Esteem Nexus 363