Sentencing Credit Laws, a Midwestern State, and Prison Discipline

Sentencing credit laws provide opportunities for inmates to gain a reduction in their prison sentence (Reference Lawrence and LyonsLawrence and Lyons 2011; Reference Weisburd and ChayetWeisburd and Chayet 1989), and such laws generally assume one of two forms—good time or earned time laws (Reference Lawrence and LyonsLawrence and Lyons 2011). Under good time laws, sentencing credits are typically awarded to inmates automatically if they follow prison rules and participate in required activities, whereas earned time laws generally only permit sentencing credits to be awarded to inmates who participate in or complete designated programs (e.g., rehabilitative treatment) (Reference LawrenceLawrence 2009). Inmates can earn good time credits in 32 states, while 37 states have laws that afford inmates earned time credits; many states permit inmates to be awarded both types of sentencing credits (Reference Lawrence and LyonsLawrence and Lyons 2011).

The Midwestern state under study here has a sentencing credit law that automatically awards inmates six months of good time for each year of their prison sentence.Footnote ¹ An iteration of the law has been in existence for over forty years, and judges in the state routinely consider the impact of the law when determining how long to sentence a criminal defendant to prison (Reference PelkaPelka et al. 2014). However, an inmate may have good time credits removed pursuant to disciplinary actions for violations of the state's inmate rules and regulations, which raises the possibility that some individuals may remain in prison longer than the sentencing judge intended.

All prisons in the state prohibit 46 acts, which are divided into three different classes of offenses—Class I offenses (e.g., assault), Class II offenses (e.g., tattoo activities), and Class III offenses (e.g., tobacco products)—that reflect their seriousness and the maximum punishment that may be imposed. Removing good time is arguably the most severe punishment prison officials can impose because inmates’ liberty interests are affected (Reference BabcockBabcock 1981). Inmates found guilty of a Class I offense in the state may lose up to two years of good time, whereas inmates found guilty of violating a Class II offense may lose up to three months of good time, and inmates found guilty of a Class III offense may lose up to two months of good time.Footnote ²

Following from the United States Supreme Court's decisions in (Wolff v. McDonnell 1974) and (Sandin v. Conner 1995), the state requires that an impartial committee conduct a disciplinary hearing in order for an inmate to lose good time. The impartial committee is comprised of senior or supervisory personnel at each prison who have received training pertaining to inmate discipline from the state's legal counsel.Footnote ³ Individuals appointed to the disciplinary committee typically serve in this capacity for several years, and some individuals serve for over a decade.

After a misconduct report is written, inmates receive notice of the hearing and the charges against them. Disciplinary hearings are expected to be held within seven days of the alleged rule violation and inmates are allowed to be present, offer evidence, and call witnesses in their defense, although the hearings are closed to the public. At the conclusion of the hearing, inmates are provided with a written statement of the decision(s) pertaining to guilt, and if relevant, the corresponding punishment(s). Aside from the maximum penalties that may be imposed for violations of different severity (e.g., Class I versus Class II), prison officials have the discretion to impose any available punishment(s) that they deem appropriate, given the circumstances pertaining to the rule violation and the inmate. Punishments for violations range from the assignment of extra work to placement in segregation. Prison officials typically remove good time in addition to imposing other sanctions (> 90 percent of cases).

Scholars who have examined the prison disciplinary process have found that inmates are rarely successful in challenging the charges against them, and despite the due process protections afforded by the Supreme Court, disciplinary committees still enjoy considerable discretion when meting out punishments for rule violations (Reference FlanaganFlanagan 1982; Harvard Center for Criminal Justice 1972; Reference HowardHoward et al. 1994; Reference ThomasThomas et al. 1991). The discretion afforded to prison disciplinary committees and the closed nature of the proceedings has the potential to generate decision-making that results in the disparate treatment of inmate groups (Reference FlanaganFlanagan 1982; Reference HowardHoward et al. 1994), though we are unware of any studies that have examined whether such treatment occurs with respect to the removal of sentencing credits. The few studies of the prison disciplinary process that exist have focused on decision-making pertaining to the use of segregation (e.g., Reference CrouchCrouch 1985; Reference FlanaganFlanagan 1982; Reference HowardHoward et al. 1994; Reference ThomasThomas et al. 1991), whereas the researchers who have assessed the effects of sentencing credit laws have examined the effects of legislative changes in these laws on offender behavior (e.g., Reference Bales and MillerBales and Miller 2012; Reference Drake, Barnoski and AosDrake, Barnoski, and Aos 2009; Reference Emshoff and DavidsonEmshoff and Davidson 1987). We attend to these gaps in the research, and contribute to the broader literature concerning decision-making related to criminal punishment, by assessing the factors that influence prison officials’ decisions to remove good time in a Midwestern state.

Influences on Prison Officials’ Punishment Decisions

Inmates charged with prison rule violations enjoy fewer rights during the disciplinary process than criminal defendants, but the punishment phase within prisons is still similar in many respects to criminal sentencing. Researchers of criminal sentencing often contrast the relevance of legal factors (e.g., offense type) versus extra-legal factors (e.g., race) (e.g., Reference FeldmeyerFeldmeyer et al. 2015; Reference JohnsonJohnson 2006; Reference KutateladzeKutateladze et al. 2014; Reference Wang and MearsWang and Mears 2010; Reference Warren, Chiricos and BalesWarren, Chiricos, and Bales 2012), and explanations why these factors affect judicial decision-making have often been framed within theories such as uncertainty avoidance (Reference AlbonettiAlbonetti 1987), causal attribution (Reference AlbonettiAlbonetti 1991; Reference Bridges and SteenBridges and Steen 1998), or focal concerns (Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998). These perspectives have also been used to understand the relative importance of legal versus extralegal factors in administrative proceedings (e.g., parole revocation) where offenders are provided fewer rights than criminal defendants (Reference Huebner and BynumHuebner and Bynum 2006; Reference Lin, Grattet and PetersiliaLin, Grattet, and Petersilia 2010; Reference SteinerSteiner et al. 2011), and so we draw on them here for our examination of prison officials’ decision-making related to good time.Footnote ⁴

The uncertainty avoidance and causal attribution perspectives on criminal sentencing are rooted in organizational theories of decision-making (e.g., Reference March and SimonMarch and Simon 1958; Reference ProttasProttas 1979; Reference SudnowSudnow 1965) and social psychological perspectives on blame attribution (Reference Carroll, Payne, Carroll and PayneCarroll and Payne 1976; Reference Farrell and HolmesFarrell and Holmes 1991; Reference HawkinsHawkins 1987). Scholars working within these frameworks have theorized that judges have an interest in reducing crime, and therefore, their decisions regarding criminal punishment rest on appraisals of offenders’ odds of recidivism (Reference AlbonettiAlbonetti 1991; Reference Bridges and SteenBridges and Steen 1998; Reference JohnsonJohnson 2006). Yet, judges typically have limited information concerning offenders’ risk of future criminality, which results in uncertainty that judges manage by relying on a bounded rationality that is the product of habit and social structure (Reference AlbonettiAlbonetti 1987, Reference Albonetti1991). That is, judges develop patterned responses to similar cases that are linked to individual and case characteristics they consider to be related to offenders’ likelihood of reoffending. The basis for their beliefs is derived from past experiences, stereotypes, and prejudices regarding particular types of cases and offenders (e.g., violent offenders, male offenders), as well as whether judges attribute blame for offenders’ criminality to personal factors such as a lack of impulse control or remorse, or to external factors such as antisocial peers or economic marginality. Offenders for whom judges perceive the cause of their criminality to be personal factors receive harsher punishment than those whose crimes are thought to be the result of environmental influences (Reference AlbonettiAlbonetti 1991; Reference Bridges and SteenBridges and Steen 1998; Reference HawkinsHawkins 1987).

The focal concerns perspective builds on the uncertainty avoidance and causal attribution frameworks by also recognizing that judges have an interest in controlling crime, but make punishment decisions in the face of uncertainty driven by limited information concerning offenders’ prospects for reform. Judges reduce the uncertainty involved in punishment decisions by developing perceptual shorthand based on experiences and stereotypes associated with individual and case characteristics (Reference AlbonettiAlbonetti 1991; Reference JohnsonJohnson 2006; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998). The focal concerns perspective further theorizes that judges are guided in their responses by three domains of reference: (1) an offender's blameworthiness, (2) an offender's risk to the community, and (3) the practical consequences of imposing the relevant punishment for an individual and/or the justice system (Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998).

Blameworthiness is associated with the retributive philosophy of punishment, such that an offender's punishment corresponds directly to their culpability for the crime and the degree of injury inflicted (Reference JohnsonJohnson 2006; Reference Steffensmeier and DemuthSteffensmeier and Demuth 2000; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998). Judges’ concerns related to offenders’ risk to the community are linked to the incapacitive and deterrent functions of punishment, and involve predictions about future dangerousness based on attributions linked to case and offender characteristics (Reference AlbonettiAlbonetti 1991; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998). Practical consequences and constraints associated with the organization and individuals also affect punishment decisions because judges are sensitive to the necessity of maintaining functional working relationships in an interdependent justice system (e.g., Reference DixonDixon 1995; Reference Eisenstein, Flemming and NardulliEisenstein, Flemming, and Nardulli 1988; Reference Ulmer and JohnsonUlmer and Johnson 2004); judges are also cognizant of the consequences of imposing punishment on particular individuals such as those with dependent children (Reference DalyDaly 1987; Reference Griffin and WooldredgeGriffin and Wooldredge 2006; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998).

Similar to judges, prison officials have an interest in controlling “institutional crime” or prison rule violations, which threaten the safety and order of a prison (Reference DiiulioDiIulio 1987; Reference HowardHoward et al. 1994; Reference Steiner and WooldredgeSteiner and Wooldredge 2009b; Reference ThomasThomas et al. 1991; Reference Toch, Adams and GrantToch, Adams, and Grant 1989). Prison officials who administer punishment in response to rule violations are also confronted by uncertainty regarding inmates’ risk to reoffend (Reference ThomasThomas et al. 1991). The relevant officials are also required to process a large volume of rule violations, while still performing other duties within the prison bureaucracy (e.g., supervisor and prison counselor) (Reference HowardHoward et al. 1994). It seems reasonable, therefore, that prison officials might also manage the uncertainty surrounding punishment decisions by developing patterned responses or perceptual shorthand linked to characteristics of rule violations and/or inmates. Prison officials may be further guided by the same focal concerns (e.g., blameworthiness) as judges.

Sentencing scholars have found that legal factors explain most of the variation in judicial sentencing decisions (e.g., Reference Feldmeyer and UlmerFeldmeyer and Ulmer 2011; Reference FeldmeyerFeldmeyer et al. 2015; Reference JohnsonJohnson 2006; Reference Johnson and DipietroJohnson and Dipietro 2012; Reference KutateladzeKutateladze et al. 2014; Reference Spohn and HolleranSpohn and Holleran 2000; Reference Ulmer and JohnsonUlmer and Johnson 2004; Reference Wang and MearsWang and Mears 2010; Reference Warren, Chiricos and BalesWarren, Chiricos, and Bales 2012; Reference WooldredgeWooldredge 2010). For prison officials, legal factors include the seriousness and type of violation, along with an inmate's violation history and security risk. Inmates who commit more serious offenses might be considered more blameworthy because offenses designated as more severe are generally those that involve a greater level of culpability and have the potential for more significant harm to the victim or the institution (e.g., assault, possession, or manufacture of dangerous contraband) (Reference FlanaganFlanagan 1982). The type of offense inmates commit may also influence prison officials’ response by shaping the categorization that officials apply to a particular case (e.g., typical drug offense). Since certain types of offenses pose a greater threat to prison safety relative to others (Reference FlanaganFlanagan 1982; Reference HowardHoward et al. 1994), the officials may perceive the inmates who commit these offenses as higher risk. Violent offenses, for instance, typically result in an injury to a person and may lead to retaliatory violence (Reference Edgar and O'DonnellEdgar and O'Donnell 1998; Reference Griffin and WooldredgeGriffin and Hepburn 2006), whereas drug or tattoo offenses often involve the use of manufactured tools that contribute to the spread of infectious diseases (e.g., Reference ClarkeClarke et al. 2001; Reference StrangStrang et al. 2006). Thus, we expect that prison officials will be more likely to remove good time credits and remove a greater amount of good time credits from inmates convicted of violations that are more serious, along with violations designated violent, tattoo, drug, or sanction violation offenses relative to other nonviolent offenses.

Prison officials may consider violations perpetrated by inmates with more significant rule violation histories or inmates designated as a higher security risk to be more the result of personal factors than environmental factors because these inmates have demonstrated a continued propensity for criminality. Inmates who have a lengthier history of rule violations (and related discipline) may also be considered more culpable, and thus, more blameworthy, because they would typically be more familiar with the prison rules and corresponding array of punishments (Reference CrouchCrouch 1985). Prison officials might also consider inmates who have a more significant violation history and those designated a higher security risk a greater risk to reoffend, since there is a link between prior criminality and offending in prison (e.g., Reference Bales and MillerBales and Miller 2012; Steiner, Butler, and Ellison 2014). We expect prison officials will remove good time more often and remove a greater number of good time credits in cases involving inmates designated a higher security risk or those involving inmates with more significant violation histories.

Given the discretion afforded to prison officials, extralegal factors could also affect their punishment decisions (Reference CrouchCrouch 1985; Reference HowardHoward et al. 1994). For instance, sentencing researchers have found that judges typically punish offenders who have more significant criminal histories and unconventional social backgrounds more severely (e.g., Reference Feldmeyer and UlmerFeldmeyer and Ulmer 2011; Reference FeldmeyerFeldmeyer et al. 2015; Reference Spohn and HolleranSpohn and Holleran 2000; Reference Wooldredge, Griffin and RauschenbergWooldredge, Griffin, and Rauschenberg 2005). Prison officials might similarly consider inmates with a more significant criminal history prior to imprisonment (i.e., inmates with a prior incarceration, those incarcerated for a violent offense) or those involved in a gang higher risk to reoffend because there is evidence of a link between prior criminality and/or gang involvement and offending in prison (e.g., Reference Bales and MillerBales and Miller 2012; Reference Griffin and WooldredgeGriffin and Hepburn 2006; Reference Steiner and WooldredgeSteiner and Wooldredge 2008, Reference Steiner and Wooldredge2009b; Reference Wooldredge, Griffin and PrattWooldredge, Griffin, and Pratt 2001). Prison officials might also consider rule violations perpetrated by these inmates to be more the result of personal factors than environmental factors because the inmates who committed them have a more significant history of involvement in deviance outside of prison. Further, inmates who have previously been imprisoned, those imprisoned for violent offenses, and gang members are labeled as such in prison, which could shape prison officials cognitive appraisals of these inmates risk to reoffend. In contrast, married inmates and those who have achieved a higher level of education may be considered less of a risk to reoffend because these inmates have demonstrated some level of commitment to conventional pursuits (e.g., Reference Wooldredge, Griffin and PrattWooldredge, Griffin, and Pratt 2001). That is, prison officials might expect inmates who have demonstrated conventional behaviors in the past to be more likely to do so in the future. We predict that prison officials will be more likely to remove good time and remove a greater amount of good time in response to rule violations committed by inmates with more significant criminal histories and inmates who have less conventional social backgrounds (e.g., gang members, unmarried, and less educated).

Inmates who are younger, male, or a member of a racial/ethnic minority group might be viewed by prison officials as higher risk to reoffend because these characteristics are overrepresented in the offender population relative to their distribution in the general population (Reference Spohn and HolleranSpohn and Holleran 2000; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998; Reference Wooldredge, Griffin and RauschenbergWooldredge, Griffin, and Rauschenberg 2005). The overrepresentation of these groups in the offender population might stimulate prison officials to categorize inmates who belong to these groups as higher risk (see Reference Wooldredge, Griffin and RauschenbergWooldredge, Griffin, and Rauschenberg 2005 for a parallel argument regarding judges). Inmates who are members of these groups (younger, males, racial/ethnic minority) may also be perceived to be higher risk because these characteristics often symbolize the “dangerous class,” which is thought to pose the greatest threat to communities (Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998). Further, there is considerable evidence to suggest that younger inmates are more likely to commit rule violations, though evidence concerning race and sex effects is mixed (e.g., Reference Bales and MillerBales and Miller 2012; Reference Griffin and WooldredgeGriffin and Hepburn 2006; Reference Harer and SteffensmeierHarer and Steffensmeier 1996; Reference Steiner and WooldredgeSteiner and Wooldredge 2009b; Reference Wooldredge, Griffin and PrattWooldredge, Griffin, and Pratt 2001). We expect that prison officials will remove good time more frequently and at a greater rate in response to violations involving inmates who are younger, male, or members of a minority group. Researchers of criminal sentencing have found that younger offenders, those who are male, or members of a minority group are typically punished more severely than their respective counterparts (e.g., Reference Feldmeyer and UlmerFeldmeyer and Ulmer 2011; Reference FeldmeyerFeldmeyer et al. 2015; Reference JohnsonJohnson 2006; Reference Johnson and DipietroJohnson and Dipietro 2012; Reference KutateladzeKutateladze et al. 2014; Reference MitchellMitchell 2005; Reference PattersonPatterson 2015; Reference Spohn and HolleranSpohn and Holleran 2000; Reference UlmerUlmer 2012; Reference Ulmer and JohnsonUlmer and Johnson 2004; Reference Warren, Chiricos and BalesWarren, Chiricos, and Bales 2012), though some researchers have found nonsignificant age and/or race effects (e.g., Reference Wooldredge, Griffin and RauschenbergWooldredge, Griffin, and Rauschenberg 2005).

Factors such as whether an inmate has children, mental health problems, or the amount of their sentence left to be served may also be relevant. Prison officials may be less likely to remove good time (or remove less good time) in cases involving inmates with children because they are aware of the practical consequences of further disrupting the tie between a parent and his or her child (see Reference Cochran and MearsCochran and Mears 2013 for a discussion of the effect of incarceration on the parent-child tie). Prison officials may also believe that offenses committed by inmates separated from their children are more the result of environmental factors than personal factors. Evidence derived from studies of judicial sentencing decisions suggests that judges typically impose less severe punishments in cases involving parents relative to nonparents (e.g., Reference Griffin and WooldredgeGriffin and Wooldredge 2006; Reference Koons-WittKoons-Witt 2002). We also predict that prison officials will be less likely to remove good time and remove fewer good time credits from inmates who have mental health problems, owing to concerns related to these individuals’ ability to do time (e.g., Reference AdamsAdams 1986; Reference FellnerFellner 2006; Reference KrelsteinKrelstein 2002), and because officials are likely to consider violations committed by these inmates attributable to environmental factors interacting with the inmates’ mental illness. Finally, we expect that inmates who have served a greater proportion of their sentence will be less likely to lose good time and lose less good time because prison officials are cognizant that the timely release of inmates has implications for levels of institutional crowding (Reference Lattimore and BakerLattimore and Baker 1992; Reference Steiner and WooldredgeSteiner and Wooldredge 2009a); changes in the release dates of inmates disrupts the consistent flow of population in and out of prisons (Reference Lattimore and BakerLattimore and Baker 1992). Institutional crowding can significantly impact prison operations, not to mention other components of the justice system (Reference JohnsonJohnson 2006; Reference Steiner and WooldredgeSteiner and Wooldredge 2009a; Reference Ulmer and JohnsonUlmer and Johnson 2004), and officials are mindful that their decisions can affect the quality of their relationships with actors across the justice system (Reference Klofas, Stojkovic and KalinichKlofas, Stojkovic, and Kalinich 1992).

Current Study and Hypotheses

As noted above, a considerable amount of research has been directed toward understanding the influences of punishment decisions leading to imprisonment; the uncertainty avoidance, causal attribution, and focal concerns perspectives have informed much of this research. In this study, we apply these perspectives to an examination of the influences of punishment decisions made within prisons, in particular, those concerning good time credits. Individuals imprisoned in the Midwestern state under study here are automatically awarded six months of good time for each year of their prison sentence, but prison officials may remove good time credits in response to violations of the inmate rules and regulations. This process is similar to a number of other states (Reference LawrenceLawrence 2009), but we know virtually nothing about the factors that influence prison officials’ decisions concerning good time. Such information is important because these decisions can affect an individual's liberty by lengthening their term of imprisonment (Reference BabcockBabcock 1981). If punishment disparities result from the decisions made by prison officials, then some offender groups may endure a greater loss of liberty relative to others and the legitimacy of the prison organization may be undermined. Drawing from theories of criminal sentencing, we examine prison officials’ decisions concerning good time. Our hypotheses derived under these perspectives (and discussed above) are contained in Table 1.

Table 1. Description of Samples of Rule Violation Incidents and Inmates Who Committed Rule Violations

^a Descriptive statistics based on N = 792 incidents.

^b Reference category.

Findings

Before delving into the results of the analysis, it is worth noting that inmates in this Midwestern state were typically convicted of rule violations that were designated as less serious offenses (i.e., Class II or Class III) and nonviolent offenses (see Table 1). Prison officials removed good time credits in response to six percent of the rule violations for which these inmates were convicted. However, 19 percent of the inmates who violated prison rules lost good time in response to at least one violation; 42 percent of the inmates lost good time in response to more than one violation. Prison officials typically removed good time in conjunction with a punishment of segregation (52 percent), followed by room restriction (21 percent).

Table 2 contains the results of the analysis of the effects of incident and inmate characteristics on prison officials’ decisions to remove good time. Officials were more likely to remove good time in response to more serious offenses (Class I and Class II) compared to less serious offenses (Class III). Officials also removed good time credits more frequently in cases involving inmates convicted of violent, tattoo, drug, and sanction violation offenses relative to cases in which inmates were found guilty of nonviolent offenses. Inmates with a more significant prior violation history, in terms of the frequency and severity of prior violations, those who had previously lost good time, and inmates who were designated a greater security risk were also more likely to lose good time. In contrast, prison officials removed good time credits less frequently in response to violations perpetrated by inmates with mental health problems or inmates who had served a greater proportion of their sentence.Footnote ⁹ Convictions for multiple violations stemming from the same incident had no effect on officials’ decision-making concerning good time. Altogether, the significant incident level characteristics explained 33 percent of the incident level variation in prison officials decisions regarding good time, and the compositional effects of the relevant incident characteristics accounted for 52 percent of the between inmate variation in the rate of good time removed.Footnote ¹⁰

Table 2. Hierarchical Bernoulli Model of Prison Officials’ Decisions to Remove Good Time Credits

Note: ^*= p ≤ .01.

Table 2 also shows that the inmate level characteristics that affected prison officials’ decision-making regarding the removal of good time included sex and gang membership. Officials were more likely to remove good time in cases involving men and in cases involving gang members. None of the other inmate level characteristics had an effect on prison officials’ decisions to remove good time, and the significant inmate level predictors only accounted for four percent of the between inmate variation in the rate of good time lost.

Turning to the results of the analysis of amount of good time lost (Table 3), it is worth reiterating that we were unable to adjust for a potential selection bias associated with the sub-sample examined here (i.e., the sample used to estimate the amount of good time removed is conditional on prison officials’ decisions to remove good time). Readers should keep this in mind when interpreting our results. It is also worth noting that we used a different criterion for determining statistical significance in the analysis of the amount of good time lost versus the analysis of whether inmates lost good time because the two analyses were based on different sample sizes.

Table 3. Hierarchical Cumulative Logit Model of Prison Officials’ Decisions Regarding Amount of Good Time Credits Removed

Note: * = p ≤ .05.

As previously noted, prison officials typically removed increments of 15 (17 percent), 30 (42 percent), 45 (23 percent), or > 45 (18 percent) days of good time. Table 3 shows that inmates lost a greater amount of good time if they were convicted of more serious offenses (Class I and Class II) compared to less serious offenses (Class III). Prison officials also removed more good time in response to drug offenses relative to other nonviolent offenses. In contrast, officials removed less good time in response to tattoo violations compared to other nonviolent offenses, while inmates convicted of violent offenses and sanction violations lost a similar amount of good time as those convicted of other nonviolent offenses. Inmates with a more significant prior violation history and those who lost good time for a prior offense had a greater amount of good time removed. Convictions for multiple violations stemming from the same incident, an inmate's security risk, their mental health problems, and the amount of their sentence served had no effect on prison officials’ decisions concerning amount of good time removed. The significant incident level characteristics explained 43 percent of the incident level variation in prison officials decision-making pertaining to the amount of good time removed, and the compositional effects of these incident characteristics accounted for 36 percent of the between inmate variation in the level of good time removed.

Table 3 also shows that the only inmate level characteristics that affected prison officials’ decision-making regarding the amount of good time removed was whether an inmate completed a GED. Officials removed less good time in cases involving inmates who had completed their GED. None of the other inmate level characteristics had an effect on prison officials’ decisions regarding the amount of good time removed, and the significant inmate level predictor did not even account for one percent of the between inmate variation in the level of good time removed.

In sum, our analyses of the factors that influence prison officials’ decisions regarding the removal of good time credits revealed that officials were more likely to consider characteristics of the rule violation rather than inmate characteristics when deciding whether to remove good time and how much good time to remove. Based on the odds ratios generated from our analyses, the strongest predictors of prison officials’ decisions related to good time included measures reflecting the seriousness (i.e., Class I) and type (i.e., violent, drug) of the rule violation, as well as an inmate's prior violation and sanction history. For instance, prison officials had 1,037 percent higher odds of removing good time in response to Class I offenses compared to Class III offenses. Inmates who were convicted of a Class I offense also had 95 percent higher odds of losing more good time than those convicted of a Class III offense. Relative to nonviolent violations (excluding drug and sanction violation offenses), prison officials had 666 percent higher odds of removing good time for violent rule violations and 587 percent higher odds of removing good time for tattoo related violations. However, inmates convicted of tattoo violations had 123 percent lower odds of losing more good time than inmates convicted of other nonviolent offenses. Each unit increase inmates accrued on the prior violation history scale was associated with a 10 percent increase in the odds prison officials removed good time and a 19 percent increase in the odds officials removed more good time. For instance, inmates who appeared before the prison disciplinary committee and had previously committed two Class II offenses and two Class I offenses had 70 percent higher odds of losing good time relative to inmates who appeared before the committee having only committed one Class I offense. Inmates who had previously lost good time had 127 percent higher odds of losing good time again, and inmates who had already lost good time had 39 percent higher odds of losing more good time than those who had not previously had good time removed.

Discussion and Conclusions

Scholars have focused considerable attention on determining whether punishment disparities result from criminal justice decision-making. Most of this research, however, has be directed toward understanding the influences of judicial decisions regarding imprisonment or parole officials decisions related to revocation or release (e.g., Reference BaumerBaumer 2013; Reference Feldmeyer and UlmerFeldmeyer and Ulmer 2011; Reference FeldmeyerFeldmeyer et al. 2015; Reference Huebner and BynumHuebner and Bynum 2006; Reference JohnsonJohnson 2006; Reference KutateladzeKutateladze et al. 2014; Reference Lin, Grattet and PetersiliaLin, Grattet, and Petersilia 2010; Reference MitchellMitchell 2005; Reference Spohn and HolleranSpohn and Holleran 2000; Reference Steffensmeier and DemuthSteffensmeier and Demuth 2000; Reference SteinerSteiner et al. 2011; Reference UlmerUlmer 2012; Reference Wang and MearsWang and Mears 2010; Reference Warren, Chiricos and BalesWarren, Chiricos, and Bales 2012; Reference WooldredgeWooldredge 2010; Reference Wooldredge, Griffin and RauschenbergWooldredge, Griffin, and Rauschenberg 2005). In this study, we examined punishment decisions made by prisons officials in response to violations of prison rules and regulations; specifically, those decisions pertaining to good time. In addition to addressing a significant gap in the literature, our inquiry is important because if punishment disparities result from these decisions, then some offender groups endure a greater loss of liberty relative to others. Disparate treatment of offender groups could undermine the legitimacy of a prison organization, which has been linked to inmates’ willingness to defy the prison rules and other legal authorities (Reference LieblingLiebling 2004; Reference Sparks, Bottoms and HaySparks, Bottoms, and Hay 1996; Reference Useem and KimballUseem and Kimball 1989).

Our analysis yielded evidence that prison officials seldom respond to rule violations by removing good time; only six percent of the rule violation incidents committed by the inmates in our sample resulted in the loss of good time, although 19 percent of the inmates convicted of a rule violation did lose good time in response to at least one violation. Following from sentencing scholars, we contrasted the relevance of legal factors (e.g., seriousness of violation) versus extra-legal factors (e.g., race/ethnicity) for predicting prison officials’ decision-making concerning good time. Our bi-level analyses provided evidence that prison officials’ decision-making regarding good time was primarily influenced by legally relevant criteria as opposed to extra-legal factors. We found that the legal factors included in our model of prison officials’ decisions to remove good time accounted for 94 percent of the explained variation at the incident-level and the compositional effects of the legally relevant factors accounted for 93 percent of the explained variation at the inmate level. We also found that the legal factors included here explained all of the explained variation at both the incident level and the inmate level (via the composition effects) in prison officials’ decisions regarding the amount of good time to remove. The strongest predictors of prison officials’ decisions to remove good time credits were legal factors reflecting the seriousness (i.e., Class I) and type (i.e., violent) of the rule violation, as well as an inmate's prior violation and sanction history.

Our expectations regarding the influences of prison officials’ decision-making concerning good time were informed by theories of criminal sentencing. We theorized that, similar to judges, prison officials reduce the uncertainty surrounding their punishment decisions by relying on patterned responses that are linked to individual and case characteristics; officials may be further guided by three domains of reference: (1) an inmate's blameworthiness, (2) an inmate's risk to the prison community, and (3) the practical consequences of imposing the relevant punishment for an inmate and/or the justice system (Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998). Aside from some important distinctions discussed below, our findings support the use of perspectives such as uncertainty avoidance, causal attribution, and focal concerns for framing potential predictors of prison officials’ punishment decisions. Data collected directly from prison officials regarding how their perceptions shape their decisions are needed, however, to assess whether they are directed by similar fields of reference (e.g., blameworthiness) as other justice system actors.

Prison officials may have removed good time more frequently and removed more good time in response to more serious violations and in response to violations perpetrated by inmates with more significant violation or sanction histories because they reasoned that the acts committed by these inmates were more attributable to personal factors rather than environmental factors. Prison officials may also have viewed inmates who committed more serious violations and inmates with longer violation or sanction histories more blameworthy. Similar to state criminal codes, offenses that are designated more severe in prison are those that involve a greater level of culpability and have the potential for more significant harm to the victim or the institution (Reference HowardHoward et al. 1994). Additionally, inmates with more extensive rule violation histories (and the corresponding discipline) could be viewed as more culpable because they are often more familiar with the prison rules and related punishments (Reference CrouchCrouch 1985). Blameworthiness is associated with the retributive philosophy of punishment, and individual culpability and the degree of harm resulting from an offense are relevant to scaling a proportional response to the offense (Reference JohnsonJohnson 2006; Reference Steffensmeier and DemuthSteffensmeier and Demuth 2000; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998).

Prisons incapacitate individuals who pose a threat to society, but prison officials also bear the responsibility for protecting the prison community (Reference ParkPark 2000). Prison officials can incapacitate an inmate for a longer period and possibly deter them from engaging in subsequent misbehavior by removing good time in response to a rule violation (Reference Weisburd and ChayetWeisburd and Chayet 1989). Based on the uncertainty avoidance, attribution, and focal concerns perspectives, officials’ decision-making concerning good time is influenced, in part, by their perceptions regarding an inmate's risk for subsequent misbehavior; these perceptions are based on attributions linked to characteristics of rule violations and inmates (Reference AlbonettiAlbonetti 1991; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998). We found evidence that prison officials removed good time more frequently in response to rule violations that would be considered more dangerous (e.g., violent, tattoo) and in response to violations committed by inmates with longer violation and sanction histories, inmates who were designated a greater security risk, male inmates, and gang members; officials remove good time less frequently in response to violations perpetrated by inmates with a higher level of education (i.e., high school diploma). We also found that officials’ removed a greater amount of good time in cases involving inmates with a longer violation and sanction history, as well as those convicted of drug offenses. In the face of uncertainty regarding an individual's prospects for reform, factors such as the nature of the offense, an individual's criminal history, and their socio-demographic characteristics often provide the foundation for justice system actors’ perceptions of high risk offenders, which ultimately shape their patterned responses to similar cases and offenders (e.g., Reference BaumerBaumer 2013; Reference Huebner and BynumHuebner and Bynum 2006; Reference Lin, Grattet and PetersiliaLin, Grattet, and Petersilia 2010; Reference Steen, Engen and GaineySteen, Engen, and Gainey 2005; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998; Reference SteinerSteiner et al. 2011; Reference Ulmer, Kurlychek and KramerUlmer, Kurlychek, and Kramer 2007; Reference Wooldredge, Griffin and RauschenbergWooldredge, Griffin, and Rauschenberg 2005). Although our findings suggest this may also be the case for prison officials, future research concerning officials’ perceptions of offenses and offenders is needed to fully understand whether these ideas apply to a prison context. Researchers might, for instance, administer surveys or conduct interviews with relevant officials to understand their perceptions regarding the cases and individuals who come to their attention, and/or to ascertain whether prison officials cognitively appraise cases and individuals in the same way as judges, prosecutors, or other justice system actors.

Prison officials’ decisions pertaining to good time may have been influenced by the practical consequences and constraints associated with the organizational environment and particular inmates. We found that officials were less likely to remove good time in response to violations committed by inmates with mental health problems and from inmates who had served a greater proportion of their sentence. Officials may have been resistant to removing good time from individuals with mental health problems due to concerns regarding these inmates’ ability to do time (e.g., Reference AdamsAdams 1986; Reference FellnerFellner 2006; Reference KrelsteinKrelstein 2002), or because they attributed blame for their acts more to the interaction between their mental illness and the environment. At the time of the study, prison crowding was also a significant problem in the state under study here, and state officials were under pressure to release inmates in a timely manner to manage levels of crowding. Institutional crowding can affect prison operations and other components of the justice system (Reference JohnsonJohnson 2006; Reference Steiner and WooldredgeSteiner and Wooldredge 2009a; Reference Ulmer and JohnsonUlmer and Johnson 2004), and it is likely that prison officials (much like judges) understand the importance of maintaining functional working relationships in an interdependent justice system (e.g., Reference DixonDixon 1995; Reference Eisenstein, Flemming and NardulliEisenstein, Flemming, and Nardulli 1988; Reference JohnsonJohnson 2006; Reference Ulmer and JohnsonUlmer and Johnson 2004). These external pressures may have influenced officials’ decisions to remove good time less often in response to violations committed by inmates who were closer to being released (i.e., inmates who had served a greater proportion of their sentence).

Although our findings generally support the application of the theories of criminal sentencing to decision-making regarding prison discipline, not all of our results were consistent with the expectations we derived under this perspective. For instance, we found that an inmate's age, race/ethnicity, and family status (marital and parental) were not relevant for shaping prison officials’ decisions pertaining to good time. We also found that an inmate's incarceration history, designation as a sexual offender, or whether they were incarcerated for a violent offense had no effect on prison officials decision-making regarding good time. These findings not only run counter to what would be expected under the uncertainty avoidance, attribution, or focal concerns perspectives, but they are also inconsistent with findings derived from studies of judicial sentencing decisions (e.g., Reference Cochran and MearsCochran and Mears 2015; Reference Griffin and WooldredgeGriffin and Wooldredge 2006; Reference JohnsonJohnson 2006; Reference Spohn and HolleranSpohn and Holleran 2000; Reference Steffensmeier and DemuthSteffensmeier and Demuth 2001; Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998; Reference Ulmer and JohnsonUlmer and Johnson 2004; Reference Ulmer, Kurlychek and KramerUlmer, Kurlychek, and Kramer 2007).

It could be that prison officials rely less on the demographic or familial characteristics of inmates than judges because prison officials typically have more information available to them to inform decision-making (Reference Gottfredson and GottfredsonGottfredson and Gottfredson 1988). For instance, prison officials typically have rich sources of information pertaining to the inmates in their custody, such as risk and need assessments, prison behavioral reports, educational test results, and progress reports related to treatment programs. The increased amount of information may have decreased the uncertainty surrounding prison officials’ punishment decisions, and permitted them to rely on more proximate indicators of an inmate's risk for subsequent misbehavior. In support of this possibility, we found that officials were influenced by an inmate's security risk, gang membership, and education, each of which is a better reflection of an inmate's risk to reoffend than their demographic or familial characteristics. Similarly, prison officials were more likely to consider an inmate's violation history and security risk than other indicators of criminal history when deciding whether to remove good time, possibly because the former would be more direct indicators of the inmate's risk for subsequent misbehavior.

An additional explanation for the relatively inconsequential effects of inmate demographic characteristics on prison officials’ decisions pertaining to good time could be the context in which this study occurred. In the close confines of a prison, the outcomes of incidents or disciplinary hearings are more widely known than those resulting from court hearings or other administrative proceedings (e.g., parole violation hearings) (Reference Bottoms and TankebeBottoms and Tankebe 2012; Reference Sparks, Bottoms and HaySparks, Bottoms, and Hay 1996). As a result, the outcomes of prison disciplinary hearings have the potential to not only affect the accused, but also the institutional environment (Reference Sparks, Bottoms and HaySparks, Bottoms, and Hay 1996). For example, if prison officials were to treat similarly situated inmates inconsistently or discriminately, then inmates may question the legitimacy of those officials and potentially become defiant (Reference LieblingLiebling 2004; Reference Sparks, Bottoms and HaySparks, Bottoms, and Hay 1996; Reference Useem and KimballUseem and Kimball 1989). Prison officials are cognizant of this situation, which may explain why the officials in this Midwestern state rely more on legally relevant criteria in making their decisions pertaining to good time than extralegal factors such as an inmate's demographic characteristics.

Taken together, the results discussed above suggest that theories of criminal sentencing are suitable to apply to an examination of punishment decision-making by prison officials, though two important caveats are worth mentioning. First, more research is needed to understand the link between prison officials’ perceptions and their decisions regarding rule violations, and whether they adopt similar fields of reference as those used by judges (e.g., blameworthiness). In particular, an ethnographic study such as the type described by (Reference Steffensmeier, Ulmer and KramerSteffensmeier, Ulmer, and Kramer 1998) might shed light on whether prison officials categorize rule violation incidents based on the characteristics of the case or the inmate, or whether they are guided by the three focal concerns during their punishment decisions in much the same was as judges or parole officials. Researchers might also survey or interview prison officials to understand their perceptions regarding rule violations and the inmates who commit them, and/or to assess whether officials view cases and individuals in the same way as other justice system actors. Second, the perspective may require minor refinement to account for the additional information typically available to prison officials relative to other justice system actors (Reference Gottfredson and GottfredsonGottfredson and Gottfredson 1988). Based on our findings, it seems that the additional information available to prison officials contributes to more rational decision-making by permitting officials to rely on more proximate indicators of risk for misbehavior (e.g., gang membership, violation history) as opposed to limited considerations of demographic or familial characteristics, and other, less proximate, indicators of criminal history (e.g., incarcerating offense). These caveats aside, however, our findings pertaining to the seriousness and type of offense, along with an inmate's violation history, security risk, sex, gang involvement, education, mental health problems, and the proportion of their sentence served were each consistent with our expectations that were derived under the uncertainty avoidance, causal attribution, and focal concerns perspectives, and suggest that these theories are useful ways of understanding punishment decisions made by prison officials.

As far as we are aware, our study is one of only a handful of studies of prison officials’ punishment decisions (e.g., Reference CrouchCrouch 1985; Reference FlanaganFlanagan 1982; Reference HowardHoward et al. 1994; Reference ThomasThomas et al. 1991), and the only study to assess punishment decision-making pertaining to sentencing credits (e.g., good time). Sentencing credit laws exist in 44 states, and these laws are often touted as a promising approach to reducing prison populations and correctional costs (Reference LawrenceLawrence 2009; Reference Lawrence and LyonsLawrence and Lyons 2011). Yet, regardless of the promise of these laws, it is important that such laws are applied fairly and equitably so as not to undermine the legitimacy of the justice system (Reference NewmanNewman 1985). Our findings suggest that for the most part, punishment decisions made by prison officials working in this Midwestern state were influenced by factors that would be considered legally relevant criteria, such as the severity of the offense, as opposed to extra-legal factors such as an inmate's race or ethnicity. However, our results are only generalizable to one state, and so additional research is sorely needed. Future research might also examine prison officials decision-making pertaining to other types of punishment (e.g., segregation), or whether sentencing credit laws achieve any of their other goals such as reductions in prison populations or inmates’ subsequent misbehavior. The need to understand the application of these laws is great, given that they exist in nearly every state (Reference Lawrence and LyonsLawrence and Lyons 2011).

In conclusion, the findings from this study suggest that prison officials in the Midwestern state under study here rarely remove good time credits in response to rule violations. In making decisions whether to remove good time and how much good time to remove, prison officials are primarily influenced by factors that would be considered legally relevant (e.g., offense severity, violation history). Extra-legal factors that affect prison officials decision-making concerning good time are typically those that are proximately connected to an inmate's risk of subsequent misbehavior (e.g., gang membership) or are those that are linked to practical consequences and constraints associated with the organizational environment (e.g., an inmate's proportion of their sentence served) and particular inmates (e.g., those with mental health problems). These findings are important because they suggest that there is very little disparity in the application of this particular punishment by prison officials, which means there is a great deal of uniformity and proportionality in punishment. However, our findings are limited to prison officials working in a single state, and given the discretion afforded to prison officials and the limited visibility associated with their punishment decisions, additional inquiry is needed to understand if our findings are generalizable to other jurisdictions. It is only by continuing to shed light on the punishment process in prisons, that we can get a better handle on how justice is administered in this unique, and often forgotten, context.