Abstract
The literature on sentencing has devoted ample consideration to how prosecutors and judges incorporate priorities such as retribution and public safety into their decision making, typically using legal and extralegal characteristics as analytic proxies. In contrast, the role of case processing efficiency in determining punishment outcomes has garnered little attention. Using recent data from a large Florida jurisdiction, we examine the influence of case screening and disposition timeliness on sentence outcomes in felony cases. We find that lengthier case processing time is highly and positively associated with punitive outcomes at sentencing. The more time prosecutors spend on a case post-filing, the more likely defendants are to receive custodial sentences and longer sentences. Case screening time, although not affecting the imposition of custodial sentences, is also positively associated with sentence length. These findings are discussed through the lens of instrumental and expressive functions of punishment.
Introduction
The stress of heavy caseloads is a part of everyday life for prosecutors (Gershowitz & Killinger, 2011). Researchers point out that modern lifestyles in industrialized countries (Brown, 2014; Quinn, 2009), the professionalization and formalization of adjudication procedures (McConville et al., 2005), and tough-on-crime policies have effected an exponential increase in cases flowing through the criminal justice funnel (Martin, 1990; Wallace, 1994). While funding for law enforcement agencies has expanded (Dunbar, 2020; Kraska & Kappeler, 1997), along with the development of new technologies to combat crime, parallel innovations in the field of prosecution are nearly nonexistent. In fact, many prosecutors still rely on paper and pen rather than electronic management systems to maintain their case files, and rarely do they examine trends across cases in their decision making (Olsen et al., 2018). At the same time, the number of cases that prosecutors must process has grown relentlessly for decades (Gershowitz & Killinger, 2011). From 1974 to 2005, the estimated number of local prosecutors in the United States only grew from about 17,000 to approximately 27,000, a 59% increase (Stuntz, 2011). However, over the same period, felony prosecutions tripled from roughly 300,000 a year to more than 1 million (Stuntz, 2011). Based on most recent estimates by the Bureau of Justice Statistics, prosecutors handle an average of 94 felony cases annually (Perry & Banks, 2011).
Without policies and practices designed to effectively manage prosecutorial resources (e.g., case prioritization and assignment policies), caseloads will continue to grow, as will delays and backlogs. On the macro policy level, coping strategies range from the decriminalization of petty offenses to expanding the use of alternatives to conviction such as diversion (Johnson et al., 2020) and deferred prosecution (Barrett et al., 2006). On the micro level, various coping strategies are reflected in individual prosecutors’ practices. One of the most influential frameworks used to explain individual criminal justice actors’ strategies to maintain work efficiency is the focal concerns perspective (Albonetti, 1986; Liu, 2020; Steffensmeier & Demuth, 2001; Steffensmeier et al., 1993; Ulmer & Johnson, 2004). From this perspective, prosecutors have limited time and a fixed scope of evidence with which to make case decisions, so they rely on stereotypes linking criminality to certain sociodemographic groups to infer the blameworthiness and dangerousness of defendants, as well as the appropriateness of particular punishments (Steffensmeier & Demuth, 2001; Ulmer & Johnson, 2004). While focal concerns theory has been criticized for being virtually untestable (Lynch, 2019), sentencing scholars have repeatedly tried to connect two of the focal concerns—defendants’ dangerousness and blameworthiness—to sentencing outcomes (Freiburger, 2009; Spohn & Beichner, 2000; Spohn & Holleran, 2000; Steffensmeier & Demuth, 2006). In contrast, far less is known about the relationship between the third focal concern—practical constraints—and its relationship with sentencing. Although “practical constraints” covers a wide range of considerations, case processing efficiency is arguably one of its main components.
It is possible that faster case processing can be harmful for defendants. Some researchers warn that prioritizing efficiency limits time to collect and examine evidence, which might compromise due process (Klein, 2003; Stuntz, 2011) and endanger sentencing fairness (Bagaric, 2015). Spending more time on a case may also result in a more punitive outcome. Recent evidence from New York County shows that prosecutors tend to make their first plea offer the most favorable, with every subsequent offer becoming more and more punitive (Kutateladze, 2018; Kutateladze et al., 2016). The rationale behind this practice might be explained through deterrence: more severe charges and longer sentences can be used as a threat for defendants who may take up more of prosecutors’ time. Conversely, defendants may get discounts when they cooperate with prosecutors and reach swift resolution in their cases. This possibility aligns with scholarship on the organizational theory, which often interprets plea negotiation as a mechanism for ensuring courtroom efficiency (Dixon, 1995; Ferrandino, 2014). Defendants who plead guilty quickly are rewarded with shorter sentences.
Lastly, it is also possible that sentence outcomes are independent of case processing times. According to the focal concerns framework, under the pressure to keep institutional efficiency, prosecutors and judges may infer defendants’ blameworthiness and dangerousness based on perceptual shorthands that assume certain social groups are more dangerous and crime-prone (Steffensmeier et al., 1993). Case efficiency, then, may function more as a reason why court actors use stereotypes to determine sanction severity, less as a direct determinant of punitiveness. In other words, efficiency might only obliquely affect sentence types and lengths, by explaining why court actors use stereotypes to determine sentences. Blameworthiness and dangerousness explain the actual variation in sentencing outcomes.
Empirical assessments of the relationship between efficiency and punitiveness remain sparse, leaving the exact nature of the relationship unclear. Yet, examining this relationship is especially timely now given that a global pandemic has triggered marked delays in case processing, with jurisdictions struggling with finding the right balance between efficiency and fairness, especially toward those in pretrial detention who cannot afford bail (Reynolds, 2020). New research and data are needed to empirically investigate how and to what extent case processing efficiency is associated with sentencing outcomes. This study directly responds to this literature void by disentangling a main question: When factors representing the other two focal concern—blameworthiness and dangerousness—are already accounted for, does case processing efficiency emerge as a meaningful predictor of (a) custodial sentences, (b) prison sentences, and (c) incarceration length? To address this question, we use fresh data from a large Florida prosecutor’s office, which include a wide range of data fields concerning offense severity, charge changes, criminal history, defense counsel type, and defendant characteristics.
Prior Research on Sentence Outcomes
The Influence of Case and Defendants Characteristics on Sentencing
Sentencing scholars have long argued that the severity of offense, prior history, and detention status are the strongest determinants of sentence type and length (Doerner & Demuth, 2010; Franklin, 2015; Johnson et al., 2008; Spohn & Welch, 1987; Ulmer, 2012). Yet various extra-legal factors such as defendant race and gender also explain sentence variations. Specifically, studies find that being young, male, and Black have been associated with more punitive outcomes at most stages of case processing, from case filing to sentencing (Ulmer & Bradley, 2017).
Offense severity is widely recognized as a key predictor of sentence outcomes (Feinberg, 1990; Kramer & Ulmer, 2009; O’Hear, 2005). According to the principle of proportionality in punishment, the severity of a sanction should increase according to the degree of offense severity. Sentencing guidelines both at the federal and state levels reflect this principle (Draper, 2009; Von Hirsch, 1992). Research also consistently finds that proportionality is upheld, using proxies ranging from the type and amount of harm done (e.g., Spohn & Spears, 1996), to statutory offense gradations (e.g., Kim et al., 2015; Kramer & Ulmer, 2009), to the number of charges and counts (e.g., Kutateladze et al., 2014; Shermer & Johnson, 2010).
Prior criminal history is another indicator that court actors use to gauge the culpability and blameworthiness of offenders. It is well-established that prior record has a direct, positive effect on sentence severity (see Spohn, 2002 for a review). Past criminal history is a factor that has been operationalized through various measures including prior arrests, prior convictions, and prior incarcerations (e.g., Spohn & Welch, 1987; Vigorita, 2001; Welch et al., 1984). Vigorita (2001) found that the count and the type of prior records had distinct effects on sentence outcomes, illustrating the importance of considering different proxies of criminal history. Researchers also find that prior offending history often explains away much of the observed extralegal disparity in sentence outcomes. For example, some researchers have found that the magnitude of racial disparities lessens considerably when the effect of prior record is controlled for (e.g., Kutateladze & Andiloro, 2014; Ulmer et al., 2016).
Pretrial detention is another strong determinant of severe sentence outcomes. Studies examining the effects of pretrial detention have generally concluded that defendants detained pretrial are more likely to receive an incarceration sentence (e.g., Dobbie et al., 2018) or a longer sentence (e.g., Oleson et al., 2017). This relationship holds even after controlling for various legal and extra-legal characteristics (Kellough & Wortley, 2002; Leiber & Fox, 2005; Phillips, 2008; Williams, 2003). This relationship may be explained through the higher likelihood to plead guilty without negotiating among detained defendants (Kutateladze et al., 2016), or it may be that detained individuals are viewed by judges and prosecutors as more dangerous (Lowenkamp et al., 2013). The influence of pretrial detention on sentencing has been documented in a multitude of jurisdictions, including New Jersey (Sacks & Ackerman, 2014), New York (Leslie & Pope, 2017), Pennsylvania (Gupta et al., 2016), Philadelphia and Miami (Dobbie et al., 2018), and Texas (Heaton et al., 2017).
Moving on to a fourth legal factor predicting sentence outcomes—mode of conviction—studies generally find strong evidence of “trial penalties” and/or “plea discounts.” Sentencing studies that assess conviction type show that, compared to pleading guilty, trial conviction increases both the likelihood and length of incarceration (Holleran & Spohn, 2004; Kurlychek & Johnson, 2004). This effect of conviction type is found across offense categories including drug crimes (Albonetti, 1997; Engen & Steen, 2000), white-collar crimes (Albonetti, 1998), violent offenses (Ulmer & Bradley, 2006), and misdemeanors (Kramer & Ulmer, 2009).
To a lesser extent, extra-legal factors also influence sentencing outcomes. Defendant demographics such as age, gender, and race may be used as perceptual shorthands by court actors to infer blameworthiness and dangerousness, leading to disadvantages in sentencing for young, minority, and male offenders (Demuth & Steffensmeier, 2004; Johnson et al., 2008). Recent sentencing studies have found that despite widespread attempts to eliminate disparity through state and federal sentencing guidelines, defendant demographic attributes still influence sentencing decisions (Franklin, 2018; Mitchell, 2005; Spohn, 2002; Ulmer, 2012). Recent scholarship explores how minority defendants experience less favorable outcomes at multiple points of case processing, which ultimately results in cumulative disadvantages expressed at sentencing (Kutateladze et al., 2014; Stolzenberg et al., 2013; Sutton, 2013).
The Importance of Case Processing Efficiency
The efficiency of case processing has long been a focus of criminal justice policymakers. Organizations such as American Bar Association and National Center for State Courts have made multiple efforts to develop standards for timely case disposition (Dodge & Pankey, 2003). For example, according to the ABA standard, 90% of felony cases should be disposed within 120 days, and 100% within 1 year (Greacen, 1988). Other think tanks have adopted similar timelines when developing performance indicators for courts and prosecutors (e.g., Measures for Justice and Prosecutorial Performance Indicators). Case processing timeliness has been echoed as an important priority by prosecutors as well (Meldrum et al., 2020). However, limited empirical research has examined whether the emphasis on case processing efficiency impacts sentencing outcomes.
At present, only a couple of studies have explicitly incorporated case processing time into analyses of sentencing outcomes. In their examination of selection bias in sentencing research, Zatz and Hagan (1985) found that the number of days between arrest and case disposition had a significant effect on sentence lengths for felony offenders in California. Interestingly, longer case processing times were associated with shorter sentences. However, more recent studies do not concur with this finding. Using federal terrorism cases, Bradley-Engen et al. (2012) found that time to conviction had a significant, positive effect on sentence length. Similarly, using court record data from Maryland, Stewart (2014) found that case processing time—the duration from filing to disposition—exerted a significant, positive effect on sentencing length.
Overall, research on case processing time and sentencing is limited, relying on old data and producing decidedly inconclusive results. However, it is important to examine the extent to which efficiency triggers punitive sentences. While saving criminal justice dollars is an important consideration, it should not be achieved at the expense of just and reasonable sentencing.
The Current Study
Using recent prosecutorial data, the present study explores two proxies of case processing timeliness—screening time and prosecution time—and their effects on punitiveness. We assess punitiveness by examining three sentencing outcomes: custodial sentences, prison sentences, and lengths of incarceration. While we acknowledge that punitiveness can take varied forms in criminal case processing, these three outcomes appear to be the most established yardsticks of punitiveness (Blumstein et al., 2005; Spohn & Spears, 1997; Tasca et al., 2019). As such, our analysis focuses on felony cases in which prosecutors decided to pursue charges and ensure convictions. While some cases are diverted and dismissed, it is unclear whether diversion and dismissal are measures of leniency, especially if these cases should not have been filed for prosecution to begin with (Kutateladze et al., 2014). Furthermore, shorter case processing time is not always better for the defendant or victim, but for felonies cases, which can languish in the system for months, speedier resolutions gain greater importance. With the growing realization that the U.S. criminal justice system is excessively punitive (Tonry, 2004), sources of punitiveness merit great attention. We turn our attention to organizational considerations in the prosecutor’s office. Sentences are not the product of the evaluation of legal factors alone; prosecutors also apply discretion that might be influenced by concerns for case processing efficiency.
Method
Data
The current study uses administrative data from the prosecutor’s office in a large Florida jurisdiction. The full dataset includes information for all felony and misdemeanor cases disposed between 2017 and 2019, capturing the progression of each case from referral through sentencing with numerous offense, offender, and case processing characteristics. We focus on assessing how, after the effects of legal factors and offender demographics are adjusted, the unexplained gaps in punishment among felony offenders are explained by the prosecutor’s time that they have consumed. 1 Misdemeanors cases as well as cases that either were not filed for prosecution or received an alternative disposition such as diversion or dismissal are excluded in the current study. This culling process resulted in an analytic sample of 14,909 cases.
Measures
Dependent variables
Three dependent variables represent distinct dimensions of the sentencing decision. The first outcome is a binary variable—custodial sentence—that captures the imposition of a custodial sentence, including jail and prison sentences. Sentences of time served only are considered noncustodial, as are probation, community service, and fines. The second outcome variable—prison sentence—looks into whether the custodial sentence includes prison versus local jail only. This variable taps into the length of the sentence—local jails are traditionally used for sentences of less than 1 year, while state prisons house defendants whose sentences are 1 year or more. However, it also speaks to distinctions in the incarceration experience. Correctional scholars note differences between jails and prisons in factors such as resources, visitation opportunities, culture, and available programming, which creates markedly different environments and experiences for the people housed in them (e.g., May et al., 2014). The third dependent variable—sentence length—is a continuous measure tapping into the number of days of a jail and/or prison sentence imposed. Less than 2% of cases had sentence length in excess of 10 years, with lengths as long as 98,915 days. Sentence length was therefore capped at 3,654 days (10 years) to avoid biased estimates. Together, these three variables provide a nuanced picture of punitiveness at the sentencing stage.
Independent Variables
Screening and prosecution time
Two independent variables capture the efficiency of prosecutorial practice. The first variable—screening time—measures the number of days that elapsed between the date the case was referred to the prosecutor’s office and the date it was filed. The second variable—prosecution time—measures the number of days that elapsed from case filing to final disposition in cases that resulted in a conviction. Unlike jurisdictions in which prosecutors are required to bring cases to arraignment quickly and typically file cases within 24 hours (e.g., see Kutateladze, & Andiloro, 2014), prosecutors in Florida have up to 30 days to make filing decisions and tend to perform more thorough case screenings (Rule 3.134 of Criminal Procedure). On average, prosecutors take an average of 24 days to file a felony case in the jurisdiction under current study. Given the time, resources, and import given to the screening decision in Florida jurisdictions, we include screening and prosecution times as separate variables in the analysis to adequately model both case processing phases in which significant prosecutorial discretion is exercised. The correlation between screening and prosecution times is .05 (two-tailed test of Pearson’s r = 0.05, p < .0001), providing additional evidence that screening and prosecution are two distinct phases of case processing.
Case and Defendant Characteristics
A range of legal and extralegal characteristics are included as controls in the analyses. To capture the severity of each felony offense, two binary variables indicate whether the top disposition charge was a first-degree felony or higher, and whether it was a second-degree felony. Third degree felonies serve as the reference category. Offense type is measured by three binary variables indicating whether the primary disposition offense is a person, property, or public order/traffic offense, with drug offenses serving as the reference category. Number of charge counts, a numeric measure of offense severity, represents the total number of disposition counts in each case. Defendants’ criminal history is measured using two count variables: number of prior arrests and number of incarcerations.
Given the clear association between pretrial detention and subsequent punishment outcomes (Oleson et al., 2016; Tartaro & Sedelmaier, 2009), detention status is included in the analyses as a binary variable indicating whether the defendant was still in pretrial detention by the time the case was disposed. Disposition type is measured as a binary variable indicating whether the case resulted in a conviction through a bench or jury trial rather than a plea. Prior research also suggests that charge changes are a key mechanism through which prosecutors influence disposition and punishment outcomes (Johnson & Larroulet, 2019), so an additional binary variable indicates whether the case’s top charge was reduced at any point between referral and disposition.
Defendant sociodemographic characteristics are also included as controls. Race/ethnicity is measured with a set of mutually exclusive binary variables showing whether the defendant is Black, Hispanic, or another race, with White defendants serving as the reference group. Gender is measured with a binary variable indicating whether the defendant is male, and defendants’ age is measured using a continuous variable. The final two variables in the analysis serve as proxies for defendants’ socioeconomic status (SES). Following past practices (Kutateladze et al., 2016; Sacks & Ackerman, 2014), we first capture SES using attorney type, which is measured as a binary variable indicating whether the defendant chose to retain a private lawyer. The second indicator of SES taps into the economic position of defendants’ neighborhoods (Rehavi & Starr, 2014). Neighborhood income is a binary variable indicating whether the defendant’s residential zip code falls into the lowest quintile of median household incomes within the study county.
Analytic Strategy
As discussed above, three distinct sentencing outcomes serve as dependent variables in the analysis. First, a series of logistic regressions examines the effects of case processing time on the odds that defendants received custodial sentences. Then, focusing on a subsample of defendants who received custodial sentences, a second set of logistic regressions assesses the effects of timing on the type of custodial sentence—jail or prison—that defendants receive. Lastly, using a series of linear regressions, we look at the association between case processing times and sentence length.
Following previous sentencing studies (e.g., Kutateladze et al., 2014; Starr, 2015; Ulmer et al., 2016), stepwise modeling was employed. The first model includes only the two independent variables—screening time and prosecution time—to enable an examination of their unconditional effects on sentencing severity. We then add case characteristics in the second model to estimate the extent to which these characteristics wash away the effects of case processing times. In the third model, we include defendant attributes while removing case characteristics. The fourth model constitutes the full model, where we include case processing times, defendant attributes, and case characteristics to estimate the independent effects of case processing times on punishment decisions.
Results
Descriptive Statistics
Table 1 provides the descriptive data for the variables employed in this study. Approximately 57.8% of this sample of felony cases result in a custodial sentence, and 26.7% of the custodial sentences are assigned to a state prison. The average custodial sentence length is 730 days. It takes an average of 24 days from referral for the prosecutor to reach a filing decision, and 82 days from filing to reach conviction. Figure 1 provides a visual comparison of the distributions of two measures of case processing time. The distribution of screening time approximates a normal distribution, with the majority of cases reaching a filing decision within 5 to 45 days. In contrast, the distribution of prosecution time displays a heavier right tail: prosecution times are positively skewed, with 25% of cases waiting at least 4 months for a disposition and about 2% of the cases remaining in the system for at least 10 months.
Descriptive Statistics (N = 14,909).
Distributions of case screening and prosecution time (days).
Two thirds of the cases involve third degree felonies (67.1%), while 27.8% are second degree felonies and just 5.1% are first degree felonies. Approximately equal percentages of cases involve person, property, and drug offenses (27.4%, 32.1%, and 28.3%, respectively), with the remaining 12.3% comprised of public order and traffic offenses. Cases involve an average of 2.1 charge counts.
Defendants have an average of 4.7 prior arrests and just 0.4 prior incarceration sentences. Over half of the defendants (55.9%) are detained at the time of case disposition. Only 1.1% of case convictions are the result of a trial rather than a plea, while 13.0% involve a charge reduction between referral and filing. The average age of defendants in the sample is 32.2 years, and most defendants are male (78.7%). The majority of defendants are Black (58.7%), while 36.9% are White and 3.5% are Hispanic. As for socioeconomic proxies, only 23.3% of the defendants retain a private attorney, and 37.3% are from a low-income neighborhood.
Predicting Custodial Sentences
Table 2 presents the results of a series of logistic regressions predicting the odds of a custodial versus noncustodial sentence. In Model 1, we find that case processing speed from filing to disposition exerts a statistically significant, positive effect on the likelihood of custodial sanctions: the odds of receiving a custodial sentence increase by 0.3% for each extra day of prosecution time (OR = 1.003, p < .001). Though small on a day-by-day basis, this effect means that when prosecutors spend 100 more days on a case post-filing, the defendant can expect 30% higher odds of receiving a custodial sentence. With regard to the effect of screening time, no statistically significant effect is found. Interestingly, including legal or extralegal controls in subsequent models does not change the effect of case processing times, which suggests that case processing time is a distinct source of punitiveness independent from the seriousness of the case or the characteristics of the defendant. In fact, supplemental analyses likewise showed that case processing did not markedly change the effect of legal variables.
Logistic Regressions Predicting Custodial Sentence Outcomes (N = 14,909).
p < .05. **p < .01. ***p < .001.
Proceeding to Model 2 (Table 2), we find that all legal factors exert a significant influence on sentence outcomes. Cases are more likely to result in custodial sentences when: they involve first or second degree felonies, the primary charge is a drug or public disorder/traffic offense, the defendant has more charges and experiences a charge reduction, the defendant has more prior arrests and prior incarcerations, the defendant is detained pretrial, and the case is disposed of through a trial. Pretrial detention emerges as the strongest predictor of a custodial sentence, increasing the likelihood sixfold (OR = 6.49, p < .001).
Model 3 (Table 2) illustrates the effects of defendants’ characteristics. Data show that age, gender, race/ethnicity, and private counseling exert a significant influence on one’s odds of receiving a custodial sentence. Men have higher odds of receiving custodial sentence than women (OR = 1.94, p < .001), and older defendants have higher odds of receiving custodial sanctions (OR = 1.03, p < .001). Although Black and White defendants have indistinguishable odds of receiving custodial sentences, being Hispanic is associated with an estimated 33% reduction in the odds of receiving a custodial sanction (OR = 0.67, p < .05). Finally, being represented by a private lawyer decreases the likelihood of custodial sentences (OR = 0.82, p < .001). The full model (Model 4 in Table 2) largely confirms these relationships.
Predicting Prison Sentences
Table 3 displays the results of four logistic models (Models 5–8) predicting the odds of prison versus jail sentences based on the subsample of defendants who received custodial sentences. We observe a similar pattern of results in this set of models. However, the influence of prosecution time is three times stronger for prison sentences outcome than custodial sentences: each one-day increase in prosecution time is associated with an estimated 1% increase in the odds of receiving a prison sentence (OR = 1.011, p < .001). This greater influence of prosecution time over prison sentences, as opposed to custodial sentences, is also reflected in the markedly larger variance explained (R2 = 0.12 compared to R2 = 0.01 in Model 1).
Logistic Regression Predicting Prison Sentence Outcome (N = 8,580).
p < .05. **p < .01. ***p < .001.
Another notable difference lies in the influence of offense severity versus detention status on the outcome of interest. Here the influence of the detention status is weaker (OR = 4.68, p < .001) while offense severity—specifically first-degree felonies—emerges as the strongest predictor of prison rather than jail sentences (OR = 8.73, p < .001). Relatedly, the effects of offense type also changed: drug crimes are now the least likely offenses to trigger prison sentences. Finally, the effect of defendant race/ethnicity is no longer a significant predictor, but the influence of gender gets stronger (OR = 2.84, p < .001).
Predicting Sentence Length
Finally, we estimate the effects of predictors on the length of incarceration among defendants who received custodial sentences (N = 8,580). Table 4 presents the results from four OLS models (Models 9–12). For the first time, both measures of case processing efficiency have significant effects. Each additional day from arrest to filing decision increases the expected sentence length by 3.6 days (p < .001), and each additional day from filing to disposition increases the length by 4.1 days (p < .001). Together, these two predictors explain a sizable portion of the variance in the dependent variable (R2 = 0.18). Similar to the results for the two previous dependent variables, nearly all legal factors exert sizable effects on sentence length. Incarceration sentences tend to be longer when: the case involves a first- or second-degree felony, the primary charge is not a drug offense, charges are reduced at some point during case processing, the defendant has more prior incarcerations, the defendant is detained pretrial, and the case is disposed of through trial as opposed to guilty plea. Note that while the trial disposition is a significant predictor of all three dependent variables, its effect is especially large for the sentence length (b = 963.17, p < .001). Another notable difference is that the number of charges and prior arrests show no statistically significant effects on sentence length, while these two matter for custodial sentences and prison sentences.
Linear Regression Predicting Custodial Sentence Length (N = 8,580).
***p < .001.
When it comes to defendant characteristics (Model 11), two predictors emerge. Sentences tend to be longer when the defendant is male (b = 242.01, p < .001) and is represented by a public defender (b = −126.73, p < .001). However, in spite of seemingly large effects of gender and attorney type, defendant characteristics explain only minimal variation in sentence length. 2 Lastly, the full model (Model 12 in Table 4) including all predictors confirms all the relationships noted above, although it does appear that the inclusion of case characteristics noticeably diminishes the effects of case processing time on sentence length. We also noticed that the effect of lawyer type is no longer significant in the full model.
Discussion
Bipartisan support for criminal justice reform is growing (The Guardian, 2019). While progressives advocate for reforms to advance justice and fairness, conservatives also embrace many of reform ideas tied to saving taxpayer dollars. Efficiency in the criminal justice system has become a particularly significant priority, especially in light of the COVID-19 pandemic, which has clogged the court system, forcing criminal justice practitioners, and policymakers to re-envision criminal case processing modes and means (Reynolds, 2020). Yet research on case processing efficiency is limited, especially in the context of prosecutorial decision making.
In this study, we examined the influence of case processing time on three proxies of punitiveness: custodial versus non-custodial sentences, prison versus jail sentences, and incarceration length. Various punishment theories and the focal concerns perspective tell us that punitiveness should be proportionate to the crime and should be reflective of the blameworthiness and dangerousness of defendants (Beccaria, 1963; Bentham, 1948), but the role of case processing efficiency is unclear in these well-established frameworks. We argue that the need for efficiency should not influence substantive justice—sentence type and length. Yet, findings from this study suggest that efficiency does affect sentencing outcomes, and that this is the case even after accounting for the influence of various other legal and extra-legal factors. Prosecution time influences all three measures of punitiveness in this study. With additional time spent on post-filing case processing, defendants are more likely to receive custodial sentences, prison sentences, and lengthy sentences. This finding aligns with some earlier work suggesting that when prosecutors spend more time on case processing, they make more punitive plea offers (e.g., Kutateladze et al., 2016).
The relationship between efficiency and punishment can be interpreted in the context of punishment as both an instrumental and expressive courtroom outcome. First, punishment may serve an instrumental purpose, insofar as the threat of less favorable future plea offers serves as a disincentive to decline early deals and prolong the case. This idea is consistent with previous research suggesting that prosecutors make the “best possible” plea offers first to head off subsequent negotiations (Kutateladze, 2018; Kutateladze et al., 2016). Thus, harsher punishment becomes an instrument to secure short case processing times.
However, the connection between prosecution time and punishment may also be due to the frustration that prosecutors experience. Feinberg defines punishment as a “conventional device for the expression of attitudes of resentment and indignation” (Feinberg, 1965, p.130). While in his writing Feinberg was contemplating broad societal condemnation of crimes, not condemnation by criminal justice actors in particular, the expressive component of crime responses can also be reasonably applied to prosecutors’ use of punishment as an emotional response to inefficient case processing. Harsher sentences in prolonged cases may be a reflection of prosecutors’ annoyance with defendants and defense counsel who consume more of their time. This “annoyance hypothesis,” to our knowledge, has not yet been tested empirically, yet it may hold the key to better explaining the relationship between efficiency concerns and punitiveness. Lending some credence to this possibility, we found that the effects of case processing times were not tied to legal or extra-legal factors, particularly for the case outcome of custodial sentence. It is possible that cases that are complex, notorious, and result in media coverage may trigger less annoyance because such cases may be prioritized by their office. Future studies should consider collecting information on media coverage and case complexity beyond charge severity to assess whether these factors mediate prosecutors’ annoyance on extended case processing time.
Unlike prosecution time, the influence of screening time was less omnipresent in the study. While screening time markedly increases the length of sentence, it has a null effect on sentence types. A post-hoc analysis shows that this pattern persists even after excluding prosecution time from the sentence type models. In an office where prosecutors file their own cases, as opposed to relying on a dedicated screening unit, prosecutors are likely to be more sensitive to how much time they spend on screening as they process the case. This finding highlights the importance of examining efficiency at various stages of case processing.
Several other controls shed additional light on the association between case processing efficiency and punitiveness, including defense counsel, detention status, and trial disposition. Nearly a quarter of the sample is represented by private lawyers, and our data show that these cases are less likely to result in custodial sentences, prison sentences, and lengthy sentences. Prior research has continually demonstrated the important role the defense counsel plays in sentence outcomes, though findings from these studies are mixed in terms of which attorney type achieves the most favorable outcomes (Ball, 2006; Cohen, 2014; Hartley et al., 2010; Kutateladze & Leimberg, 2019).
We also identity a very strong and consistent effect of pretrial detention on sentencing, which reinforce the findings of previous studies on this topic (Oleson et al., 2016; Tartaro & Sedelmaier, 2009; Williams, 2003). Nearly 56% of our full sample were detained at disposition, and our data show that defendants who are detained pretrial are substantially more likely to receive punitive sentencing outcomes. Even among those charged with a low-level (third-degree) felony, about half of the defendants were in pretrial detention, suggesting that the ability to afford bail plays a large role in determining detention and ultimately sentencing outcomes. Several newly elected prosecutors are daring to confront the questions of bail reform—even advocating to end the use of cash bail (Iannelli, 2020; Powers, 2020)—yet it remains unclear to what extent election campaigns translate into actual policies once these reform-minded prosecutors take office and face pressures to increase efficiency and keep cases moving (Richardson & Kutateladze, 2020).
Lastly, scholars have repeatedly pointed to a relationship between guilty plea/trial disposition and sentencing (McCoy, 2005; Ulmer & Bradley, 2006). As Abrams (2013) writes in his description of the trial penalty: “longer sentences are necessary in order to induce settlement and without a high settlement rate, it would be impossible for courts as currently structured to sustain their immense caseload” (p. 777). Our data show that the practical constraints of handling immense caseloads make plea disposition a go-to option in most cases, as only 1% of cases resulted in trial disposition. The effect of a trial disposition, though rare, is so consistent that trial disposition emerged as a significant predictor of all three sentencing outcomes. Individuals who opt for trial are noticeably more likely to be incarcerated, receive prison sentences, and face longer sentences. While in this particular jurisdiction, there are no written plea guidelines, it appears that guilty pleas are rewarded with favorable treatment.
Although the present study helps to advance the literature on sentencing, punitiveness, and prosecution efficiency with new data, we would like to acknowledge several limitations of this analysis. The first limitation concerns the generalizability of our findings to other prosecutorial offices, especially those outside Florida. Therefore, multi-jurisdictional and comparative studies are desirable next steps to advance this line of work. Second, the present study focuses on the direct effect of case processing time on punishment; however, it is possible that it also interacts with other predictors to indirectly exert influence on sanctions. Researchers should also test whether case efficiency interacts with legal and extra-legal factors such as defendant race and offense type to indirectly exert influence on sanction outcomes. Another direction for future research could be assessing how prosecutors’ caseloads, level of experience, race, and gender may mediate the relationship between case processing time and punishment. Lastly, due to data limitation, we do not have sufficient data to gauge prosecutors’ rationale to deliver more punitive sanctions to defendants who consumed more of their time. Mixed-method and qualitative research has great potential to better explain prosecutorial concerns for efficiency, as well as the pathways through which these concerns translate into greater punitiveness. Future efforts that focus on understanding how prosecutors think about case processing, prioritize efficiency as a goal in prosecution, and make plea decisions will do much to move research on the field of prosecution forward.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the John D. and Catherine T. MacArthur Foundation under grant G-1706-152065.
