Abstract
Drawing on importation and deprivation perspectives and integrating developmental/life-course criminology, the current study examines predictors of change in major and minor misconduct incidences, over a 16-year period, among a sample of 1,125 adults in custody serving long-term sentences. Findings from mixed-effects regression models revealed that factors of the prison environment such as custody level, visitation, facility transfers, and program participation directly shape the behavior of those incarcerated. Most notably, findings suggest that factors that explain differences in misconduct among those incarcerated may not necessarily explain individual changes in misconduct. Results highlight the important need for research that focuses on the experiences of individuals with long-term sentences, and the value of integrating a life-course perspective to inform policy that extends beyond static risk assessments to understand the influence of correctional factors that may promote behavior change.
Introduction
Research in developmental and life-course criminology has traditionally focused on criminal behavior before incarceration and after release, emphasizing how age-graded transitions and turning points, such as employment, marriage, or military services, shape persistence and desistance from crime over time (Akers et al., 2020; Laub & Sampson, 2001). Within this framework, incarceration is often treated as a discrete event that interrupts the life-course and potentially alters future offending trajectories. Although this line of research has considerably advanced our understanding of criminal behaviors, scholars have noted that it often overlooks periods of incarceration or conceptualizes imprisonment as a time when offending simply stops (Eggleston et al., 2004). As a result, far less attention has been paid to incarceration itself as a prolonged life-course context. For many individuals, imprisonment can span decades and encompass substantial portions of the adult life-course. In fact, over half (56%) of those in prison are serving a sentence of 10 years or more (The Sentencing Project, 2022), while 16% of those in U.S. prisons are serving a life, or nearly life, sentence (The Sentencing Project, 2025).
For individuals serving long-term sentences, incarceration is not a brief interruption but a sustained social environment in which behavior continues to develop, stabilize, or change over time. Within this context, institutional misconduct represents a critical yet understudied form of criminal behavior within life-course criminology, reflecting continued engagement in rule-violating behavior while confined. A growing body of research has begun to examine misconduct through developmental and life-course perspectives, demonstrating that misconduct is not static and individuals follow heterogeneous behavioral patterns during incarceration (Buckner & Jennings, 2025; Cihan, Davidson, & Sorensen, 2017; Cihan et al., 2020; Cihan, Sorensen, & Chism, 2017; Cihan & Sorensen, 2019; Cochran & Mears, 2017; DeLisi, 2003; Silver & Nedelec, 2018). Incarceration presents a unique environment, where deciding to desist in criminal behavior is contingent on personal factors as well as the structural setting of prison (Lugo, 2020). The prison environment can allow individuals to sever ties with anti-social peers and provide a rigid structure with opportunities for rehabilitation, new job skills, and parenting techniques that promote desistance and long-term behavioral change (Lugo, 2020; Sampson & Laub, 2003). However, the prison experience can also promote the continuation of crime. Prison can introduce individuals to anti-social peers, which allows for continued social learning of criminal behavior, while simultaneously diminishing social bonds and inducing strain through limited opportunity and harsh conditions (Nagin et al., 2009; Steiner et al., 2014). However, this literature remains constrained by cross-sectional designs, relatively short observation periods, or analytic approaches that emphasize differences between individuals rather than changes within the individual over time (Butler et al., 2022; Day et al., 2015). As a result, less is known about how the evolving conditions and experiences of incarceration shape within-person changes in misconduct across extended periods of confinement, limiting understanding of prison as a dynamic life-course context.
With the growth in population of individuals serving long-term sentences in U.S. prisons, prison constitutes a time-structured life-course context in which institutional conditions, relationships, and opportunities change over time. Building on life-course criminology and research on institutional misconduct, the current study conceptualizes incarceration as a dynamic social environment and examines predictors of both between-individual differences and within-individual change in major and minor misconduct over time. Using 16 years of longitudinal data from adults serving long-term sentences, this study integrates importation, deprivation, and management perspectives to identify institutional and individual factors associated with behavioral change during imprisonment, thereby extending life-course theory into the prison context.
A Life-Course Framework for Institutional Misconduct
Research examining correlates of institutional misconduct has largely been guided by three interrelated theoretical perspectives: importation theory, deprivation theory, and management (Steiner, 2018). Although these perspectives are often discussed together, they emphasize distinct mechanisms through which misconduct may emerge and change over time. Together, these perspectives can be used to highlight incarceration as a prolonged, time-structured life-course context in which both enduring conditions of confinement and dynamic institutional practices shape behavior over time.
The importation model attributes misconduct to individual characteristics and values that individuals “import” into the prison setting, including demographic characteristics, criminal histories, and prior socialization experiences and are expressed through destructive behavior, disruptive attitudes, and cultural values (Cressey & Irwin, 1962; Steiner et al., 2014). From this perspective, misconduct reflects the continuation of a broader criminal subculture within institutional settings, where the “inmate code” mirrors the “street code” or norms learned outside of prison (Cressey & Irwin, 1962). Research applying an importation perspective have identified factors such as entry age, race/ethnicity, sex, prior criminal history, education, gang membership, mental health, prior employment, and the offense for incarceration as variables “imported” into the prison setting that impact misconduct (Cunningham & Sorensen, 2006, 2007; Cunningham et al., 2005; DeLisi, 2003; Griffin & Hepburn, 2006; Morris et al., 2012). Importantly, these factors are largely time-invariant, making them well-suited for explaining differences between individuals but less informative for understanding behavior change during incarceration.
In contrast, the deprivation model emphasizes structural conditions of confinement and the strains produced by the harsh prison environment, including the loss of autonomy, restricted movement, and limited social contact that exerts physical and mental deprivations on individuals (Clemmer, 1940; Steiner et al., 2014; Sykes, 1958). The deprivation perspective conceptualizes misconduct as an adaptive response to these conditions, which may vary in salience across the incarceration period. Research drawing on this perspective has linked misconduct to factors such as prior prison misconduct (Meyers et al., 2018; Quick et al., 2023; Salerno & Zgoba, 2020; Steiner & Cain, 2016), prison crowding (Glazener & Nakamura, 2020), access to programs and employment while incarcerated (Duwe & McNeeley, 2020; French & Gendreau, 2006; Pompoco et al., 2017), misconduct sanctions (Steiner & Cain, 2019), and visitation (Cihan et al., 2020; Reidy & Sorensen, 2020; Turanovic & Tasca, 2017). However, deprivation-based studies have often treated these conditions as static or uniform, limiting insight into how their effects may evolve over time.
Management perspectives further distinguish misconduct as a product of institutional governance, organizational practices, and policy implementation, emphasizing how correctional decision-making structures opportunity and constraint within the prison setting. The framework posits that weak prison management, excessive control, rapid changes to policy and procedure, and other characteristics of the prison setting can impact behavior (Hausam et al., 2020; Steiner et al., 2014). Unlike deprivation, which reflects the enduring conditions of confinement, management focuses on dynamic institutional processes such as custody classification, transfers, staffing levels, procedural justice, and program allocation (Beijersbergen et al., 2015; Steiner et al., 2014). These factors are inherently time-varying and may function as institutional turning points that redirect behavior during incarceration.
Findings from research investigating the relationship of these factors and misconduct have largely remained consistent (Steiner et al., 2014). At the same time, however, studies using longitudinal designs and advanced analytic approaches have produced contradictory findings in some cases (Cihan & Sorensen, 2019; Glazener & Nakamura, 2020; Hilinski-Rosick & Freiburger, 2016; Kigerl & Hamilton, 2016; Linning et al., 2022; Silver & Nedelec, 2018). This research has revealed the heterogeneity in misconduct trajectories and underscored the importance of temporal ordering, but has often focused on identifying trajectory groups rather than examining within-individual change.
Time-Varying Predictors of Misconduct
Age and Misconduct
Age has been the most consistent and significant predictor of misconduct. Specifically, prior research shows an inverse relationship, where those that are younger are more likely to engage in misconduct (e.g., Cunningham et al., 2005; Flanagan, 1980; Gonçalves et al., 2014; Kigerl & Hamilton, 2016; Steiner et al., 2014). However, age is typically included as a control variable, which does not account for contextual factors of the age and misconduct relationship (Augustyn et al., 2020; Blowers & Blevins, 2015). Recent studies employing longitudinal analysis such as group-based trajectory modeling suggest that the relationship between age and misconduct may be more complex. For example, several studies have found individuals who display delayed-onset or escalating patterns of misconduct over time (Cihan, Davidson, & Sorensen, 2017; Cihan & Sorensen, 2019; Cochran & Mears, 2017; Morris et al., 2012; Reidy et al., 2017, 2018).
Custody Level
As a management tool, custody level classifications serve to maintain order and safety, such that those who pose the greatest risk for violence and disorder are subject to the greatest restrictions. While structure and organization can differ, correctional facilities have hierarchical structures that can include both external and internal custody levels. External custody level refers to the characteristics of the prison and its level of perimeter security (Cihan & Sorensen, 2019). Internal custody level refers to the assigned custody that dictates individual freedom within the facility such as limited access to move about, the work assignment and programs available, and housing assignments (Cihan & Sorensen, 2019). Studies examining custody level and misconduct have found that higher custody levels are associated with increases in misconduct (Bosma et al., 2020; Glazener & Nakamura, 2020; Gover et al., 2008; Logan et al., 2023; Sorensen & Davis, 2011). Similar conclusions have been produced in longitudinal studies, where those with higher custody levels are more likely to be in persistent or chronic misconduct groups (Cihan, Davidson, & Sorensen, 2017; Cihan et al., 2020; Cihan & Sorensen, 2019; Kigerl & Hamilton, 2016; Sorensen & Reidy, 2019).
Despite consistent findings in the relationship between custody level and misconduct, studies have been limited in their operationalization of custody level. As mentioned above, most studies include custody level without distinguishing between prison or individual custody level. This is a crucial distinction because an adult in custody may be housed in a maximum-security prison, but their assigned classification or custody level will impact their daily patterns and freedom. This is an important consideration, and when overlooked, it fails to account for individual changes in custody level over time as well as the effect that custody level can have on other predictors of misconduct, such as visitation and other privileges (Worrall & Morris, 2011).
Work and Programming
In a meta-analysis of 68 studies, French and Gendreau (2006) found that vocational and educational programs largely had no significant impact on misconduct. However, some research suggests an inverse relationship where involvement in these programs reduces instances of misconduct (Logan et al., 2023; Meyers et al., 2018). For example, Steiner and Wooldredge (2008) found that the number of hours spent per week in work assignments reduced several forms of misconduct, and Logan and colleagues (2023) found that the number of prison programs that individuals started reduced total and violent misconduct. In a recent study, Duwe and McNeely (2020) found that while participation in prison labor industries was associated with a significant increase in misconduct and time to first misconduct, the percentage of time spent in prison industries significantly decreased the odds of misconduct and a slower time to the first misconduct violation. Linning and colleagues (2022) examined the effects of programming over a 3-year period using a cross-lagged panel analysis. They found that overall, misconduct declined during the 3-year study period and during program participation, however, rates were higher before and after exposure to programming. Linning and colleagues (2022) suggest that this may be attributed to a “backfire” effect.
Time Served
The amount of time spent incarcerated is a deprivation itself, given the restrictions and privileges that are accessible to the general public (Sykes, 1958). The amount of time served in prison has been examined as an adjustment process where individuals learn the formal and informal rules that govern prison interactions and living (Bottoms, 1999). Prior research has not specified the amount of time served but suggests that adjustment to prison changes the longer one is incarcerated, where factors and strains of the prison environment may be more pronounced when first entering (Quick et al., 2023). For example, Kreager and colleagues (2017) found that when compared to younger adults in custody serving shorter sentences, those who have spent more time incarcerated had a vested interest in an orderly environment. Conversely, research has also suggested that spending more time incarcerated can induce further strain and frustration that leads to adverse behavioral effects and violent misconduct (Berg & DeLisi, 2006; Nagin et al., 2009). Findings from research, however, largely support the former. Specifically, those who have spent more time incarcerated are less likely to engage in violent misconduct (Sorensen & Davis, 2011; Steiner & Cain, 2019) as well as misconduct in general (Toman, 2022). This has also held true in studies that examine trajectories of misconduct, where most individuals decline over time (Cihan, Davidson, & Sorensen, 2017; Cihan et al., 2020; Cihan & Sorensen, 2019; Silver & Nedelec, 2018).
Visitation
Regarding visitation, empirical studies have found both desistance-promoting and criminogenic effects. For example, Turanovic and Tasca (2017) found that visitation led to feelings of support, love, and comfort for some, while others felt sad, guilt, and stress. Conversely, research has shown that early and consistent visitation leads to lower levels of misconduct (Cihan et al., 2020; Cochran, 2012). Thus, visitation may help cope with an individual’s feelings of alienation and poor conditions that are experienced while incarcerated, but it may also lead to feelings of hopelessness and depression. Siennick and colleagues (2013) examined the effects of visitation on misconduct in the 6 weeks leading up to visitation and 6 weeks after. Findings indicated that misconduct gradually declined in the 3 weeks leading up to visitation and were 48% lower than baseline right before visits, but then sharply increased in the 4 weeks after, before returning to baseline at 6 weeks (Siennick et al., 2013).
While the prior research discussed above has considerably expanded knowledge on the correlates of correctional misconduct, limitations remain. Specifically, most prior work has been cross-sectional in design (Butler et al., 2022; DeLisi et al., 2011; Steiner et al., 2014). When considering recent longitudinal work, most have been limited to short time periods, and typically examine a specific institutional measure such as a program (Linning et al., 2022), visitation (Cihan et al., 2020; Siennick et al., 2013), crowding (Glazener & Nakamura, 2020), procedural justice (Beijersbergen et al., 2015) or disciplinary confinement (Picon et al., 2022; Toman, 2022). In addition, while several studies examining trajectories of misconduct have emerged recently (Cihan, Davidson, & Sorensen, 2017; Cihan et al., 2020; Cihan, Sorensen, & Chism, 2017; Cihan & Sorensen, 2019; Cochran, 2012; Reidy et al., 2018), these do not account for individual change over time, but rather those following similar patterns of behavior.
Current Study
With this in mind, the purpose of the current study is to examine how time-varying prison experiences shape change in institutional misconduct among adults serving long-term sentences. Specifically, we ask whether factors commonly associated with misconduct, such as custody level, visitation, transfers, and program participation, explain differences between individuals and changes within individuals over time. Without consideration of facility-level factors, research may overestimate the variance in correctional misconduct that is attributed to pre- prison/importation predictors. Much of the prior research implicitly assumes that predictors of misconduct operate similarly across individuals and within individuals, yet life-course theory suggests that stability and change may be driven by different processes (Laub & Sampson, 2001). Factors that distinguish individuals with higher overall misconduct may not necessarily explain why a given person’s behavior improves or worsens at points during incarceration.
Using 16 years of longitudinal data and mixed-effects models, this study seeks to advance the literature by explicitly separating between-person and within-person effects. In doing so, we extend life-course criminology into the prison context and provide a framework for identifying institutional conditions that may promote behavioral change during long-term confinement. The current study will examine misconduct records for a period of 16 years, incorporating both time-variant and time-invariant predictors into models. This research seeks to contribute to the knowledge on the correlates of misconduct and the factors that lead to decreases and increases in this behavior, to better inform policy, correctional staff on how to improve prison practices, and provide scholars with new avenues of consideration.
Data and Methods
Sample
Data were provided by the Department of Corrections in a Northwestern state in the United States. Electronic files were obtained on all adults in custody who were currently incarcerated, and had served at least 15 consecutive years at the point when data were provided in March 2023. This resulted in 1,179 adults in custody who initially entered prison at any point between 1974 and 2008. Given the limited number of adults in custody who were female (N = 47) and serving a sentence for a non-violent crime (N = 7), the study was restricted to males convicted of violent offenses (N = 1,125). Since participants entered prison at various time points, the analysis was restricted to 16 years to ensure data were available for the complete sample (i.e., 16 waves). In addition, since electronic disciplinary records were only available post-1996, a control measure for the number of years incarcerated was included to account for those who entered prison prior to 1996.
Dependent Variable
Three measures of misconduct were used in the current study. Total misconduct was measured as the total number of disciplinary incidents recorded for each adult in custody, each year over the 16-year period (
Time-Invariant Measures
Importation factors consistent with prior research that were included in the current study were prior incarceration history, number of current offenses, sentence length, and demographic characteristics (Cunningham & Sorensen, 2007; Griffin & Hepburn, 2006; Siennick et al., 2013; Steiner et al., 2014; Toman, 2022). Race/ethnicity (0 = White, 1 = Non-White) and life sentence (0 = no, 1 = death, life with parole, life without) were included as binary measures. Age at entry, number of current offenses, years incarcerated, and number of prior incarcerations 2 were measured continuously. Sentence length was measured continuously and capped at 80 years to avoid skewness (15–230 years) and to include those without an anticipated release date (i.e., death, life with and without parole). Finally, Wave measures time from the initial wave (year) and Wave2 accounts for the possible non-linearity in misconduct over time.
Time-Variant Measures
Custody level was measured as the highest recorded custody level at each wave. Custody levels range from one to five (
Analytic Strategy
To examine how factors relate to changes in misconduct over time and to control for salient time-variant and invariant variables, a mixed-effects longitudinal model was used. Mixed-effects models have become a popular form of longitudinal analysis, as they combine fixed- and random effects models to allow for greater flexibility in models (Allison, 2005; Firebaugh et al., 2013). Mixed-effects longitudinal analysis presents a random intercept for each individual that allows for variation across time. Using this method, time is nested within respondents, while the model also estimates between-individual effects (Allison, 2005). Then, within-individual change is estimated as the deviation from the individual mean at each time point. This approach results in the same estimates provided by fixed-effects model along with estimations for time-invariant predictors (Rabe-Hesketh & Skrondal, 2012). To capture temporal ordering, time-lagged models were used, where time-variant factors are measured as t – 1 to ensure that all time-variant covariates occurred prior to misconduct (Kaiser & Reisig, 2019).
Results
Tables 1, 3, and 4 present the linear mixed-effects models to examine the effects of time-varying prison factors on correctional misconduct, net of time-stable controls. When looking at all forms of correctional misconduct (Table 2), transfers, visitation, custody level, and program participation were all significantly associated with between-individual misconduct. This indicated that individuals who experienced more transfers (b = 1.62, p = .000), had a higher custody level (b = .99, p = .000), and participated in prison programming (b = .35, p = .01) all had higher levels of misconduct compared to others. Those who received more visitation (b = −.04, p = .000) had less misconduct than those with lower visitation. When looking at the within-individual effects, different findings emerge. Program participation (b = .10, p = .000) was positive and significantly related to within-individual change in misconduct, while individual-level changes in being transferred (b = −.06, p = .00) more often resulted in lower levels of misconduct. However, visitation and custody level were null, meaning that changes in an individual’s level of visitation and custody did not show a statistically significant change in levels of misconduct.
Descriptive Statistics for Study Variables (N = 1,125, NT = 18,000)
Note. SD = standard deviation; NT = nested observations.
Mixed-Effects Regression Estimates Predicting Misconduct (Lagged Time-Varying Covariates)
Note. Values are unstandardized regression coefficients (b) and standard errors (SEs).
p < .05. **p < .01. ***p < .001.
Mixed-Effects Regression Estimates Predicting Major Misconduct (Lagged Time-Varying Covariates)
Note. Values are unstandardized regression coefficients (b) and standard errors (SEs).
p < .05. **p < .01. ***p < .001.
Mixed-Effects Regression Estimates Predicting Minor Misconduct (Lagged Time-Varying Covariates)
Note. Values are unstandardized regression coefficients (b) and standard errors (SEs).
p < .05. **p < .01. ***p < .001.
Turning to major forms of misconduct violations (Table 3), when looking at the between-individual effects, individuals that experienced more transfers (b = .96, p = .000), had a higher custody level (b = .57, p = .000), and participated in more programs (b = .16, p = .04) were associated with more major misconduct events than other individuals. Individuals who received more visitation (b = −.02, p = .000) had lower instances of major misconduct than others. When looking at within-individual effects, prison programming (b = .05, p = .01) was significant and positively related to changes in major misconduct; however, within-individual changes in the number of transfers (b = −.06, p = .000) were negatively associated with changes in major misconduct. This indicated that when individuals were transferred more often over time, there was an associated decrease in major misconduct incidents. Similar to model one, custody level, visitation, and job participation were not related to within-individual major misconduct.
Finally, when looking at minor misconduct (Table 4), findings were generally consistent with model one. Transfers, visitation, custody level, and program participation were all significantly associated with between-individual differences in minor misconduct. This indicated that individuals who were transferred (b = .64, p = .000) to other facilities more often, had higher custody levels (b = .39, p = .000), and participated in programming (b = .16, p = .01), were more likely to engage in minor misconduct over time, while those that received more visitation (b = −.02, p = .000) were less likely to engage in minor misconduct than those with less visitation. When looking at within-individual effects, program participation (b = .05, p = .001) was positive and statistically significant in within-individual change in levels of minor misconduct, while transfers (b = .00, p = .74), visitation (b = −.00, p = .64), custody level (b = .01, p = .30), and job participation (b = −.00, p = .75) were not significant in changes if minor misconduct.
Discussion
The goal of the current study was to broaden developmental life-course theory and corrections research by investigating the factors influencing changes in misconduct over time among individuals serving long-term sentences. Most importantly, findings showed that experiences within the prison environment directly shape the behavior of those incarcerated, through the use of longitudinal regression models that revealed between- and within-individual effects. Results from the current study support prior research that has used cross-sectional designs and shorter time frames, while others provide new insights to expand the knowledge base. Three major findings are discussed below.
First, those with higher custody levels were more likely to engage in misconduct and those who received more visitations were less likely to engage in misconduct, regardless of the type of infraction. However, this relationship was only significant between individuals. This falls in line with prior research (Cihan et al., 2020; Glazener & Nakamura, 2020; Kigerl & Hamilton, 2016; Siennick et al., 2013; Sorensen & Davis, 2011), while also providing additional context to relationships. For example, prior research, both cross-sectional (Bosma et al., 2020; Gover et al., 2008; Sorensen & Davis, 2011) and longitudinal (Cihan, Davidson, & Sorensen, 2017; Cihan et al., 2020; Cihan & Sorensen, 2019; Sorensen & Reidy, 2019), has typically included a static measure of custody level at entry without distinguishing between the facility or individual custody level. While the current findings provide further support for the causal influence of custody level status on subsequent behavior, they also suggest that this relationship may be more complex than previously thought. Specifically, custody level was not statistically significant for within-individual changes of misconduct. While this may be due to little deviation in custody level (within-individual) relative to their mean (between-individual) over time, the deprivation theory may also lend a possible explanation.
The deprivation perspective explains that prison is a deprivation itself, given the restricting environment that limits privileges (Sykes, 1958), and that the amount of time spent incarcerated includes an adjustment process where individuals learn the official and unofficial rules that regulate interactions and living (Bottoms, 1999). For those who have spent a considerable amount of time incarcerated, the change in custody level that is accompanied by changes to movement restrictions and privileges may be less pronounced, or may have a vested concern in maintaining an orderly environment (Kreager et al., 2017).
Second, findings suggest that receiving visitation is important to reducing misconduct between individuals, but that the frequency at which one receives visitation over time may not influence behavior. While there has not been much research to longitudinally examine the relationship between visitation and misconduct over this length of time, some prior research may support this finding. For example, Siennick and colleagues (2013) examined the effects of visitation on misconduct over 17 months and found that misconduct gradually declined in the 3 weeks leading up to visitation, but then sharply increased in the 4 weeks after, before returning to baseline at 6 weeks (Siennick et al., 2013). Thus, it may be the case that any effects of visitation on misconduct are acute, which would support the null findings of any within-individual change over time. Simply put, this may suggest that receiving visitation throughout incarceration reduces the risk of engaging in misconduct, but the amount of visitation does not.
Finally, the findings from transfers and programming were less consistent and perhaps suggest a more complex relationship with misconduct. Being transferred to another facility increased misconduct regardless of type; however, the within-individual effects varied. Specifically, being transferred more often was associated with a significant decrease in major misconduct incidents and increased minor misconduct incidents, but was not a significant predictor of all misconduct incidents. When explaining this finding, it is important to note that the reason for being transferred was not included. Therefore, it may be the case that being transferred to another facility for a serious/major misconduct incident reduces subsequent incidents, because of the separation of those involved. It is also possible that being transferred to a more restrictive prison limits the opportunity to engage in further misconduct. Future research would benefit from considering the reason for being transferred as well as the duration spent at each facility to better understand this relationship. Similarly, participating in prison programming was associated with a significant between-individual increase in all misconduct incidents and minor incidents, but not related to major misconduct specifically, while there was a significant increase in within-individual misconduct regardless of type. While prior scholarship examining the relationship between programming/work and misconduct is mixed, similar findings have been noted. For example, Linning and colleagues (2022) examined the effects of programming over a 3-year period and found that misconduct declined overall during the study period; however, rates were higher before and after exposure to programming. Linning and colleagues (2022) suggested that this may be attributed to a “backfire” effect. Furthermore, Logan and colleagues (2023) found that the top 1% of those most involved in misconduct participated in fewer prison programs. This finding may also illustrate that prison programs may simply represent an opportunity for misconduct by increasing their interactions with and access to others within the facility. This may be an avenue for future exploration on the effects of prison programming on institutional misconduct. Still, it is important to note that both program and job participation were measured dichotomously, therefore, the dosage effect could not be assessed. Future research should include a measure of the time spent in programs to better capture the anticipated benefits programs are presumed to have.
These findings not only expand our understanding of the prison experience and how factors of the prison environment are associated with misconduct, but also underscore the importance and utility of applying a developmental life-course perspective to the study of institutional behavior. A central goal of life-course research is to explain both stability and change in behavior over time and to identify the characteristics and contexts present at different stages of life that promote the persistence or desistence in behavior (Elder, 1994). By extending this perspective to incarceration, the current study findings suggest that prison itself may function as a time-structured life-course context in which individuals serving long-term sentences may experience distinct stages of prison life. Through this lens, factors of the prison environment are not static predictors of misconduct but rather may be contingent upon the individual’s experience and/or stage of incarceration. Their influence may depend on cumulative exposure to incarceration, institutional adaptation, and the timing of specific experiences or interventions. The distinction of between-individual differences and within-individual change observed in this study highlights that some institutional conditions may explain overall levels of misconduct, whereas others may operate as dynamic mechanisms capable of shaping behavioral change during incarceration. Together, these findings suggest that applying a life-course perspective to prison misconduct research offers a valuable framework for understanding how institutional environments actively structure behavior over extended periods of confinement.
While the current study provides new insights into the relationship between crucial factors of the prison environment/experience and misconduct, it is not without limitations. First, the sample for the current study came from a single state and consisted of males imprisoned for violent offenses. While this is to be expected given the focus of those with long-term sentences, it does limit the generalizability of results. Furthermore, the sample primarily consisted of individuals reporting a race/ethnicity of white, which limited the ability to further examine racial differences. While this is consistent with the racial and ethnic representation within the current studies state, future research should consider using larger, regional/national samples that include female adults in custody. Second, due to the frequency at which individuals were transferred, observations could not be nested within facilities. Meaning, facility characteristics that may impact misconduct and bias covariate estimates such as visitation procedures, availability of jobs and programs, and factors such as overcrowding and inmate-to-staff ratio, were not included. To further capture this, future research would benefit from multi-level mixed-effects models that allow for nesting within a facility. Several known correlates of misconduct were also not available in the current dataset. This includes factors of the importation perspective such as education, marital status, and gang affiliation, as well as developmental life-course predictors such as social and offending history, and mental health factors. Finally, it is important to consider that the most serious infraction was recorded when examining models for different types of misconduct. The finding that results were generally consistent across models may be because individuals engaging in more serious misconduct infractions are also engaging in technical infractions. Future research would benefit from including measures for the number and type of infractions that occur within each incident to further understand how factors differentially impact various types of behavior.
Notwithstanding the above-mentioned limitations, findings from the current study can provide useful new understanding and consideration for policy, practitioners, and scholars to address misconduct. Most notably, these findings indicate that factors that may explain differences in misconduct among individuals incarcerated may not necessarily be the same factors that influence individual changes in misconduct. With that in mind, more research should consider isolating which institutional factors may best promote individual desistance from misconduct. Finally, this study highlights an important need for research to focus on the trajectories and experiences of individuals who experience long-term incarceration. These findings reinforce the value of a life-course lens for uncovering the dynamic processes that shape misconduct in prisons. Using a life-course perspective to examine patterns of misconduct over the long term can inform strategies that move beyond static risk assessments to recognizing the impacts of environmental factors that may actively promote behavioral change over time.
Footnotes
Authors’ Note:
There are no conflicts of interest or funding to report.
