What happens to imprisonment rates when a progressive prosecutor is elected: Quasi-experimental evidence from Cook County

Abstract

There has been a clear recognition in recent years that the United States' extremely high rates of imprisonment, coupled with significant racial disparities, are problematic. The progressive prosecutor movement has developed in response to these issues and has gained increasing attention over the past decade. Despite the growing importance of the progressive prosecutor movement in criminal legal reform efforts, we know little about the extent to which electing a progressive chief prosecutor actually leads to reductions in overall imprisonment rates and racial disparities in imprisonment, with no studies to our knowledge using data that capture preelection differences in imprisonment trends across jurisdictions. In this study, we begin to fill this gap by examining the implications of the 2016 election of Kim Foxx in Cook County, Illinois, for overall, Black, White, and Latinx imprisonment rates. Employing a quasi-experimental synthetic control approach that accounts for preelection differences in imprisonment trends, our findings suggest that Kim Foxx's election led to a reduction in the imprisonment rate overall, as well as for Black and White individuals. However, our findings suggest that substantial racial disparities in imprisonment persisted after her election.

Keywords

progressive prosecution mass imprisonment racial disparities criminal legal reform

Introduction

In the past 50 years, the criminal legal system in the United States has been characterized by its punitive policy and wide racial disparities. The United States incarcerates people at a rate that is more than four times the world average, at 629 people incarcerated for every 100,000 people (Fair and Walmsley, 2021). Black and Latinx individuals disproportionately bear the burden of this punitiveness, experiencing imprisonment rates that are nearly 5 times and approximately 1.3 times the rate of White imprisonment rates, respectively (Nellis, 2021). This approach to punishment has made the United States an outlier among advanced democratic countries, and it comes with serious consequences (Gottlieb, 2017; Wildeman, 2016). Incarceration has negative social, economic, and health implications, both for the person incarcerated and that person's family and community (Clear, 2009; Epperson and Pettus-Davis, 2017; Wakefield and Uggen, 2010).

There has been a clear recognition in recent years that the United States’ extremely high rates of imprisonment, coupled with significant racial disparities, are problematic (Enns, 2014; Gottlieb, 2022; Gottschalk, 2015). The progressive prosecutor movement has developed in response to these issues and has gained increasing attention over the past decade (Bazelon, 2020). Drawing on the fact that chief prosecutors are elected (often in low-turnout elections), this movement has sought to elect progressive prosecutors that run on platforms of reducing overall imprisonment, as well as racial disparities in the prison system, through the use of the substantial discretion that they have with charging decisions (Davis, 2019). While traditional prosecutors certainly remain far more common than progressive ones, progressive prosecutors are increasingly winning election (Morrison, 2020; Pickerell, 2020).

Despite the growing importance of the progressive prosecutor movement in criminal legal reform efforts, we know little about the extent to which electing a progressive chief prosecutor actually leads to reductions in imprisonment rates overall and racial disparities in imprisonment (see Mitchell et al., 2022, for an exception). In fact, to our knowledge, no studies have sought to answer these questions with data that have allowed researchers to account for differences in imprisonment trends between jurisdictions with and without a progressive prosecutor prior to the election of the progressive chief prosecutor. As such, any observed differences found in prior work may be due to preexisting county differences, rather than the election of the progressive prosecutor.

In this study, we begin to fill this gap by using the 2016 election of Kim Foxx (one of the first and most prominent progressive prosecutors in the country) in Cook County, Illinois, as a quantitative case study. While there is no clear definition of a progressive prosecutor, Kim Foxx clearly fits the bill, as she campaigned around progressive priorities, has been referred to as a progressive prosecutor by the media, criminal justice organizations, and in academic articles, and has enacted policies that are progressive given current criminal legal reform standards (Davis, 2019; Levin, 2021; Pickerell, 2020). Drawing on data from 2006 to 2019, we employ a synthetic control approach to create a control group that consists of a weighted combination of Illinois counties that most closely match Cook County on trends in the outcomes of interest prior to Kim Foxx's election, but that did not elect an individual known to be a progressive chief prosecutor during the study period. Our findings suggest that Kim Foxx's election led to a reduction in the imprisonment rate overall, as well as for Black and White individuals. However, we also find that racial disparities in imprisonment remained substantial in the aftermath of her election.

Prosecutors and the buildup of mass incarceration

Prosecutors are unique actors in the criminal legal system. They have large amounts of discretion and hold the responsibility of deciding whether and when to bring charges, and what those charges contain (Bazelon, 2020; Kreag, 2017; Stuntz, 2006). The political nature of prosecutors’ offices can lead to perverse incentives. Prosecutorial success is difficult to measure, and thus the number of convictions achieved is often used as a proxy (Kreag, 2017). Since conviction rates do not distinguish between plea bargaining and jury conviction, prosecutors, when looking to increase their conviction success rate, turn to strategies such as offering generous plea deals or trying to pressure defendants into taking pleas (Kreag, 2017). Because punitive metrics are primarily used to judge a prosecutor's performance, prosecutors are often incentivized to send as many people to prison as possible (Bazelon, 2020; Pfaff, 2017).

Proponents of the view that prosecutors are a primary contributor to mass incarceration argue that the increased aggression of prosecutors in pursuing charges has been responsible for the growth in imprisonment (Pfaff, 2017). From this perspective, as prosecutors try to show the public that they are serious about safety, they pursue heavier charges, encouraging plea deals that often lead to lengthy prison sentences (Pfaff, 2017). Moreover, the creation of mandatory sentencing laws gave prosecutors more power and made their decision on what charge, if any, to bring all the more important—because if sentenced, the judge would have to grant at least a certain number of years in prison (Lynch, 2023; Starr and Rehavi, 2013).

This is not to say that there is a consensus on the power that prosecutors have in the criminal legal system. Some have argued that the District Attorney's power has been exaggerated in attempts to find solutions to prison buildups (Bellin, 2019). The power that prosecutors have often does depend on the other actors in the system working with them for their desired outcome, rather than pushing against their aims (Bellin, 2019; Fryer, 2020). This argument points out how prosecutors are dependent on the police to investigate crimes, legislators to make laws that the prosecutor must uphold, and judges to actually sentence someone to prison (Fryer, 2020). Therefore, disagreement centers less on whether prosecutors matter and more on how much they matter.

This debate on the extent to which prosecutors matter has not been settled by empirical analyses. On one hand, Pfaff (2012) provides empirical evidence that he argues suggests that, at least since 1994, increases in imprisonment have largely been driven by prosecutors. Specifically, Pfaff claims that prosecutors significantly increased how frequently they filed charges against arrestees, while arrests, arrests per crime, prison admissions per felony filing, and time served have not increased (Pfaff, 2012). However, a number of scholars have argued that Pfaff empirically overstates the importance of prosecutors by relying on flawed measurement and data, with findings driven at least in part by a flawed measure of time served and changing state court reporting practices (Beckett, 2018; Bellin, 2018). Additionally, recent work by Neal and Rick (2023) suggests that policy matters most for imprisonment growth and that sentencing and parole policies are likely most responsible for mass incarceration. While debate about the relative importance of the role prosecutors played in incarceration growth remains contested, scholarship has documented that there is significant variation across jurisdictions in the charging decisions prosecutors make and characteristics of prosecutor's office, such as political affiliation, shape incarceration outcomes (e.g., Arora, 2018; Johnson, 2018; Lynch, Barno and Omori, 2021). Taken together, it is certainly plausible (although not guaranteed) that electing a reform-minded progressive prosecutor could lead to a reduction in the imprisonment rate.

The progressive prosecutor movement

While prosecutors’ considerable power and discretion have remained largely unchanged over time, there has been a notable shift in some prosecutors’ offices over the past decade (Davis, 2019). This shift, which has come to be termed “progressive prosecution,” often includes an acknowledgement of the harms of primarily punitive approaches, as well as a recognition that these approaches may be ineffective at promoting safety or justice (Davis, 2019). Although not a monolith, “progressive” prosecutors center criminal legal reform as a means to uphold public safety, by pledging to use their power to reduce the use of incarceration and to make the criminal legal system fairer and more equitable (Davis, 2019). Some of the policies and practices associated with progressive prosecutors include: declining to prosecute classes of cases deemed less serious, creating and expanding programs to divert individuals away from traditional prosecution, recalibrating sentencing recommendations, and holding police accountable for unethical or illegal behavior (Davis, 2019; Sklansky, 2017). A common goal espoused by progressive prosecutors is to address the long-standing racial inequities in the criminal legal system through prosecutorial reforms (Fryer, 2020; Levin, 2021). Indeed, Fair and Just Prosecution, a national network of progressive prosecutors, names addressing racial disparities as one of the core principles for the twenty-first century prosecutor, stating that prosecutors should “publicly commit to reducing racial and ethnic disparities that arise from prosecutorial practices” (Fair and Just Prosecution, p. 15).

Developing alongside the progressive prosecutor movement's momentum has been a range of backlash, critique, and recommendations for the movement to better achieve its stated goals. For instance, one of the most visible progressive prosecutors, Larry Krasner, in Philadelphia, has endured extensive criticism for implementing policies such as refusal to prosecute marijuana possession and prostitution charges, and a commitment to prosecute police corruption and excessive violence (McGraw, 2021). These stances have garnered Krasner considerable opposition, which include Philadelphia's police union and a political action committee formed by retired police that opposed Krasner's reelection, and the Pennsylvania House of Representatives recently impeached him (although the Senate has yet to hold a trial) (Nichols, 2021; Schultz, 2023). A key element undergirding this opposition is the concern that electing a progressive prosecutor will lead to significant increases in violent crime and homicide (Hogan, 2022; Petersen et al., 2024).

Other critiques of the progressive prosecutor movement caution against assuming that systemic change can be driven by one sector of the criminal legal system, particularly addressing the systemic racism and inequities in incarceration and punishment (Fryer, 2020). Gajwani and Lesser (2019) suggest that existing progressive prosecutor approaches will be insufficient to enact system change and address violent crime, and that restorative justice practices may help achieve these goals. An additional limitation to the overall effect of progressive prosecutors is that, despite their growth, progressive prosecutors still comprise a relatively small minority of the more than 2300 prosecutors’ offices in the U.S. (Bureau of Justice Statistics, n.d.)

A central question at the heart of the promise of the progressive prosecutor movement and its criticism is the question of effectiveness. Because progressive prosecutors are a fairly recent phenomenon, there is little extant research to address the question of effectiveness, and the studies that do exist can be broken into two buckets: (1) Those that focus on the implications for violent crime and homicide and (2) The extent to which progressive prosecutors are taking steps toward the systems change that they seek by reducing the scale and increasing the equity of the criminal legal system.

We are aware of two published studies that have assessed the implications of electing a progressive prosecutor for crime and violence. In the first, Hogan (2022) finds that deprosecution in Philadelphia from 2015 to 2019 led to a significant increase in homicides compared to a synthetic control group of other large cities. In the second, Petersen and colleagues (2024) use a difference in difference approach to assess the implications of electing a progressive prosecutor for county-level crime rates in the 100 most populous U.S. counties. The results from this analysis differed from Hogan (2022) in that the authors found little evidence that electing a progressive prosecutor led to increased violent crime (although they did find evidence of a roughly 7% increase in property crime).

Our study does not focus on crime and instead focuses on whether elected progressive prosecutors meet their aims of reducing the scale and increasing the equity of the criminal legal system. We are aware of two studies that directly assess this. One study suggests that expectations should be tempered: while reform-minded priorities may be present in progressive prosecutor jurisdictions, there can be challenges with implementation, such as overcoming existing norms and ambiguity in policy (Richardson and Kutateladze, 2021). However, in the only quantitative study we are aware of that directly examines case outcomes associated with progressive chief prosecutors, Mitchell et al. (2022) analyzed a random sample of Florida felony cases, finding that cases in jurisdictions with a progressive prosecutor were less likely to produce a felony conviction and had lower levels of racial disparities in felony convictions than jurisdictions with more traditional prosecutor orientations.

Although the Mitchell et al. (2022) study is novel and advances our understanding of the implications of progressive prosecution, a key limitation is that the analyses focused on cases from a single year and do not capture the effects of changing from a traditional to a progressive prosecutor. For instance, it is quite possible that the communities in Florida that had progressive prosecutors at the time of the study were different in many ways than the communities that did not, even before a progressive prosecutor was elected. If this is the case, these community level differences that preceded the election of the progressive prosecutor may explain any differences in criminal case outcomes between progressive and traditional jurisdictions. To address this concern and better estimate the causal effects of progressive prosecution, quasi-experimental analyses that leverage jurisdictional changes from a traditional to a progressive prosecutor are necessary. Thus far, no studies to our knowledge have taken this approach when assessing how progressive prosecutors influence the scale and equity of the criminal legal system.

Cook County progressive prosecutor context

Chicago and Cook County have been at the center of the building progressive prosecutor movement since its origins. As the largest county in the state of Illinois, containing more than five million residents, Cook County has been the focus of much discussion over criminal legal practices (Bocanegra, 2019). While Cook County has a history of a substantial carceral state, it has in recent years also become a central player in the progressive prosecutor movement. In 2016, Cook County elected Kim Foxx as State's Attorney, ousting the incumbent, Anita Alvarez, who had focused her time in office on aggressive prosecution and protecting police officers from misconduct accusations (Davis, 2019). This close relationship with the police, and tough on crime emphasis, was more closely aligned with the traditional prosecutor's office. Yet, it was this very relationship with the police that contributed to Alvarez's reelection loss, after she was slow to prosecute the police officers who shot and killed Laquan McDonald (Neyfakh, 2015).

Kim Foxx, however, ran on a progressive platform, and after her election, has worked to enact policies designed to achieve criminal legal reform. One of the first policies Foxx implemented was to raise the threshold for felony retail theft from $300 to $1000 (Schmadeke, 2016), which Bourne (2024) found led to significant reduction in felony theft caseloads, jail bookings, prison admissions, and custodial sentencing time (with relative benefits often greatest to White defendants) among the targeted group (people charged with theft). Soon after the felony theft threshold policy change, Foxx instructed prosecutors to release anyone held on a bond of $1000 or less and announced a policy recommending more proactive use of recognizance bonds (Reclaim Chicago et al., n.d.). In 2018, to increase transparency and accountability at the Cook County State's Attorney Office, Foxx's office released six years of felony criminal case data for use by researchers and policy makers (Rice, 2018). The office also more actively advocated for line prosecutors to increase their use of diversion and other alternative prosecutions. As a result, in Foxx's first two years in office, 25% more people facing felony charges were diverted compared to the number of people diverted under Anita Alvarez (Reclaim Chicago et al., 2021). At the same time, Foxx's office focused on law enforcement of violent offenses, including establishing the office's first gun crimes strategies unit (Ali, 2020). Therefore, Foxx's efforts to reduce imprisonment focused on being less punitive around nonviolent offenses, rather than changing punishment responses to violence.

The current study

In the current study, we begin to fill this gap examining the effect of Kim Foxx's election on Cook County's: (1) Overall imprisonment rates; (2) White imprisonment rates; (3) Black imprisonment rates; and (4) Latinx imprisonment rates. Based on the literature discussed above, we anticipate that her election will be associated with reductions in Cook County's imprisonment rate overall and for each racial/ethnic group. While imprisonment disparities by race exist for nonviolent offenses, imprisonment disparities are actually greater for violent offenses (Carson, 2018; Sabol and Johnson, 2022). Since Kim Foxx's reforms have largely focused on nonviolent offenses, we do not expect her election to disproportionately benefit people of color.

Data, measures, and analytic strategy

Data

The data for our analyses come from two sources. First, we capture county-level imprisonment data from the Illinois Department of Corrections (IDOC) Prison Population Data Sets (IDOC, n.d.). Specifically, we combined each end-of-year prison population dataset to create a panel that captures the number of people imprisoned in Illinois at the end of each year for the years 2006–2019. We use information on the county where an individual was sentenced to prison to determine the number of people that were imprisoned at the end of a given year for each county. This data also indicate the race/ethnicity of each person imprisoned, enabling us to determine the number of White, Black, and Latinx individuals who were imprisoned at the end of a given year for each county. Since the population of counties vary, these raw totals do not accurately capture an individual's risk of experiencing imprisonment. Therefore, we use bridged race population estimates from CDC Wonder for each county to create imprisonment rates that account for county differences in population (CDC, n.d.).

Measures

For our analyses, we rely on four measures: (1) The overall imprisonment rate; (2) The White imprisonment rate; (3) The Black imprisonment rate; and (4) The Latinx imprisonment rate. The overall imprisonment rate in a given county in a given year was captured using this equation: OverallPrisonRate_yct= ((PrisonIncar_yct)_/(Pop_yct))*(100,000), with PrisonIncar_yct representing the number of people in prison in a county in a given year and Pop_yct being an indicator of the size of a county's population in a given year. Therefore, this overall imprisonment rate number is equal to the total number of people in a given county and year that were incarcerated in state prison per 100,000 population. We did not include individuals in Federal prisons because Federal prison numbers should only be minimally impacted (if at all) by actions of county-level decisionmakers.

Then, we used the same formula to create race-specific imprisonment rates, except overall prison and population numbers were replaced by corresponding numbers that were specific to each racial/ethnic group. The formula for the White imprisonment rate is: WhitePrisonRate_yct= ((WhitePrisonIncar_yct)/ (WhitePop_yct))*(100,000). The Black imprisonment rate is captured as follows: BlackPrisonRate_yct= ((BlackPrisonIncar_yct)/ (BlackPop_yct))*(100,000). Last, the Latinx imprisonment rate is measured as: LatinxPrisonRate_yct= ((LatinxPrisonIncar_yct)/ (LatinxPop_yct))*(100,000). Therefore, these race-specific imprisonment rates capture the total number of White (Black) (Latinx) people in a given county and year that were incarcerated in state prison per 100,000 White (Black) (Latinx) population.

Analytic strategy

To examine the impact of Kim Foxx's election on Cook County's imprisonment rate overall, and among Black, White, and Latinx individuals, we treated Cook County as a quantitative case study and employed a quasi-experimental synthetic control approach (Abadie, Diamond and Hainmueller, 2015; Bartos and Kubrin, 2018; Gottlieb, Harper and Jung, 2024). The synthetic control approach is an extension of the difference in difference framework, which relies on the assumption that trends in the outcome of interest are parallel across the treatment and control groups and would continue to be so if the treatment was not implemented. Unfortunately, this assumption is often implausible, since it can be challenging to identify a location or group of locations with average trends that are parallel to the treated location (Bartos and Kubrin, 2018).

Unlike the difference in difference approach, the synthetic control method does not rely on this parallel trends assumption. Instead, it weights potential donor (control) locations to identify a control group that most closely matches the treated group on pretreatment trends in the outcome of interest using a data-driven approach (time invariant, non-negative, and add to one; Abadie et al., 2015; Bartos and Kubrin, 2018). At this juncture, two types of covariates can be included: (1) prior outcome values and (2) predictors (other than the outcome) that are known to be causally associated with the outcome (Bartos et al., 2020; McCleary et al., 2017; McLelland and Mucciolo, 2022). In this case, we are not certain what the causal predictors are for imprisonment rates overall, as well as for race-specific imprisonment rates. Additionally, because rates of imprisonment are impacted both by prior and current policies and trends, they tend to change slowly over time, unless there is a significant shock to the criminal legal system. As such, prior values of the outcome are most likely to be the strongest predictor of current values, so we include all pretreatment values of the outcome as covariates; an added benefit of this approach is that it maximizes pretreatment fit (Ferman, Pinto and Possebom, 2020; McCleary et al., 2017). When all pretreatment outcome values are included as covariates, it is recommended to not include other covariates, since they have no impact on the selection of the synthetic control (Ferman et al., 2020; McLelland and Mucciolo, 2022). We, therefore, do not include covariates other than the pretreatment values of the outcome of interest. In instances where pretreatment trends in the outcome of the treated group closely match pretreatment trends of the control group, posttreatment differences between the treated and control group are viewed as an estimate of the causal effect because without the intervention the treated group would be expected to follow the posttreatment outcome path of the control group (which may or may not be the same as its own pretreatment trend).

For situations where the data-driven approach is not able to create a weighted control group that closely matches the pretreatment trends of the treatment group, bias-corrected synthetic control approaches have been developed. These approaches use regression to estimate the amount of bias introduced into synthetic control estimates from bad pretreatment fit and then de-bias the traditional synthetic control estimate with the regression estimate (Abadie and L’Hour, 2021; Ben-Michael et al., 2021; Wiltshire, 2022). There are not explicit guidelines to determine whether pretreatment fit is good, but we employ the following requirements for overall, Black, and Latinx rates: (1) In the year before the treatment, the control group differs by less than 1% from the treatment group and (2) The average difference between the control and treated groups is less than 1% throughout the pretreatment period. Because White imprisonment rates tend to be lower (in Cook County they were less than 100 per 100,000 population every year), 1% differences can be quite small; therefore, for White rates we applied the following thresholds: (1) In the year before the treatment, the control group differs by less than 1 per 100,000 population from the treatment group and (2) The average difference during the pretreatment period between the control and treated groups is less than 1 per 100,000 population.

An important step before conducting a synthetic control analysis is to identify appropriate potential control counties. In our case, we began by restricting the potential controls to counties in Illinois, which has the benefit of controlling for state policy changes that have the potential to significantly impact imprisonment rates. A key point to highlight is that the control group is not constructed based on population characteristics (e.g., size of the county), for which no set of Illinois counties would be a good match. Instead, because our interest is in rates of imprisonment, which adjust for the size of the population, it is much more likely that some combination of Illinois counties will be similar to Cook on pretreatment trends in these population adjusted rates. In addition to restricting our sample to Illinois counties, further sample restrictions were based on the need to avoid overfitting (which introduces bias and is increasingly likely as the number of potential control counties increases), while also maintaining an appropriate number of potential controls to conduct statistical significance tests (Abadie, 2021).

We adopted a systematic approach to identifying potential donor counties. We began by running each analysis with all Illinois counties (other than Cook), potentially serving as donor counties. However, for each of these analyses, results using Stata's allsynth command indicated that overfitting was an issue (Wiltshire, 2022). Next, we restricted our sample to counties that had a sizable number of people at risk for imprisonment (i.e., a sizable number of people overall for the total imprisonment rate analysis, a sizeable number of Black people for the Black imprisonment rate analysis, a sizeable number of White people for the White imprisonment rate analysis, and a sizable number of Latinx individuals for the Latinx imprisonment rate analysis). In our first cut, we restricted our sample to counties that had populations above the median pooled sample value in each year of the study (26,288 people for the total imprisonment analysis; 24,151 White people for the White imprisonment analysis; 962 Black people for the Black imprisonment analysis; and 608 Latinx people for Latinx imprisonment analyses).

After these restrictions, analyses using Stata's allsynth command indicated that overfitting was no longer an issue for overall imprisonment rates and White imprisonment rates, but remained so for Black and Latinx imprisonment rates (Wiltshire, 2022). Therefore, we further restricted these two samples to counties that had populations above the 75th percentile pooled sample value (i.e., had at least 3032 Black people for the Black imprisonment rate analysis and had at least 2646 Latinx individuals for the Latinx imprisonment rate analysis). After these restrictions, overfitting was no longer an issue for the Black and Latinx imprisonment rate analyses (Wiltshire, 2022). Therefore, our final samples consisted of the following: (1) Total Imprisonment Rate: All counties with at least 26,288 people in each year; (2) White imprisonment rate: All counties with at least 24,151 White individuals in each year; (3) Black imprisonment rate: All counties with at least 3032 Black individuals in each year; and (4) Latinx imprisonment rate: All counties with at least 2646 Latinx individuals in each year.

After establishing the potential pool of donor counties for each analysis, we conducted four separate analyses. The pretreatment period began in 2006 (the first year in which data were available) and ended in 2016 (Kim Foxx began her term at the start of 2017). For each analysis, we matched Cook County to a synthetic control that most closely matched pretreatment trends in the imprisonment rate outcome of interest. For overall imprisonment rates, Black imprisonment rates, and White imprisonment rates, the pretreatment fit was good (the control met our threshold on average during the pretreatment period, as well as in the last pretreatment year), so we used the traditional synthetic control approach for these analyses. For Latinx imprisonment rates, the pretreatment fit was not as good: the control was 1.47% different from Cook County on average during the pretreatment period and 1.55% different in the year prior to the treatment. We, therefore, employ a bias-corrected synthetic control approach for that analysis, using OLS regression to determine the amount of bias introduced by poor pretreatment fit and then using that estimate to de-bias the results (Wiltshire, 2022).

Determining whether associations are statistically significant is not as straightforward in the synthetic control approach as it is in traditional regression approaches (Abadie et al., 2015). Rather than relying on coefficients and standard errors, the synthetic control method determines statistical significance using placebo-in-space estimates (Abadie et al., 2015; Bartos and Kubrin, 2018; Galiani and Quistorff, 2017). In short, we created a synthetic control group for each potential control county in each analysis and act as if each potential control county was treated at the same time as the treated unit. Then, we rank each county by the size of its effect, the ratio of the posttreatment root mean squared prediction error (RMSPE) over the pretreatment RMSPE (Galiani and Quistorff, 2017; Wiltshire, 2022). The effects of Kim Foxx's election were deemed to be statistically significant if at least 95% of the control units have smaller posttreatment effects than Cook County (Galiani and Quistorff, 2017). Each analysis was conducted in Stata 17 (StataCorp, 2021).

Results

Overall imprisonment rates

Our first analysis examines the association between the election of Kim Foxx and Cook County's imprisonment rate. To do so, we constructed a synthetic control unit with imprisonment rates most similar to those in Cook County from 2006 to 2016. Figure 1 shows the imprisonment rate in Cook County (solid black line) and the synthetic control unit (dashed black line) pre- and post-election. Moreover, Figure 1 also documents the counties that make up the synthetic control and how much weight each is given. As the figure shows, the pretreatment differences between Cook County and its synthetic control were quite small (the synthetic control differed by 0.39% from Cook County on average during the pretreatment period and 0.18% in the year prior to treatment). Therefore, we treat any posttreatment differences in overall imprisonment rates between Cook County and its synthetic control as the effect of Kim Foxx's election on imprisonment rates overall.

As Figure 1 illustrates, beginning in 2014–2015 (before Kim Foxx was elected), the overall imprisonment rate began declining in both Cook County and its synthetic control. However, after Kim Foxx's election, Cook County's imprisonment rate continued to decline while the imprisonment rate in the control group declined slightly in 2017 before increasing slightly in 2018 and 2019. Therefore, Kim Foxx's election appears to be associated with substantively significant reductions in Cook County's overall imprisonment rate, particularly in 2018 and 2019. Specifically, by 2018, the synthetic control's imprisonment rate was 40.43 per 100,000 population (11.19%) higher than Cook County's. By 2019, the gap grew further: the synthetic control's imprisonment rate was 69.13 per 100,000 population (20.65%) higher than Cook County's. Placebo synthetic control estimates for the other 50 counties in this sample suggest that the associations in 2018 and 2019 were not due to chance (p = 0.02 in both instances; for 2017, p = 0.20).

White imprisonment rates

Next, we assessed the impact of Kim Foxx's election on race-specific imprisonment rates, beginning with White imprisonment rates. To do so, we constructed a synthetic control that had White imprisonment rates most similar to those in Cook County from 2006 to 2016. Figure 2 shows the White imprisonment rate in Cook County (solid black line) and the synthetic control (dashed black line) pre- and post-election and also documents the composition of the synthetic control. As Figure 2 demonstrates, pretreatment differences in the White incarceration rate between Cook County and its synthetic control were small (the synthetic control differed by 0.52 per 100,000 people from Cook County on average during the pretreatment period and 0.57 per 100,000 population in the year prior to the treatment). Therefore, we treat differences between Cook County and its synthetic control posttreatment as the effect of Kim Foxx's election on White imprisonment rates.

As was the case with overall imprisonment rates, Figure 2 shows that the White imprisonment rate began declining from 2014 to 2015 (before Kim Foxx was elected) in both Cook County and its synthetic control. Once Kim Foxx was elected, Cook County's White imprisonment rate continued to decline while the White imprisonment rate in the control group declined slightly in 2017 before starting to increase by a small amount in 2018 and 2019. Therefore, Kim Foxx's election appears to be associated with a substantively important reduction in the White imprisonment rate in Cook County. Specifically, in the first year (2017), the synthetic control's White imprisonment rate was 5.60 per 100,000 population (7.28%) higher than in Cook County. The magnitude of the association grew in the next two years, with the synthetic control having a White imprisonment rate that was 12.51 per 100,000 population (17.39%) and 22.17 per 100,000 population (32.65%) higher than the White imprisonment rate in Cook County in 2018 and 2019, respectively. Placebo synthetic control estimates for the other 50 counties in this sample suggest that the associations in 2017, 2018, and 2019 were not due to chance (p = 0.02 in 2017; p = 0.001 in 2018; and p = 0.02 in 2019).

Black imprisonment rates

In our third analysis, we examined the implications of Kim Foxx's election on Black imprisonment rates. To do so, we constructed a synthetic control with Black imprisonment rates that most closely matched those in Cook County from 2006 to 2016. Figure 3 illustrates trends in the Black imprisonment rate in Cook County (solid black line) and the synthetic control (dashed black line) pre- and post-election and also documents the composition of the synthetic control. As Figure 3 shows, Cook County and its synthetic control had very similar pretreatment trends in Black imprisonment rates (the synthetic control differed by 0.15% from Cook County on average during the pretreatment period and 0.11% in the year prior to the treatment). Therefore, we treat posttreatment differences between Cook County and its synthetic control as the effect of Kim Foxx's election on Black imprisonment rates.

As was the case with the overall imprisonment rate and White imprisonment rate, Figure 3 illustrates that beginning in 2014–2015 (before Kim Foxx was elected) the Black imprisonment rate began declining in both Cook County and its synthetic control. However, after Kim Foxx's election, Cook County's Black imprisonment rate continued to decline while the Black imprisonment rate in the control group declined slightly in 2017 and then increased slightly in 2018 and 2019. As is evident from Figure 3, the impact of Kim Foxx's election on the Black imprisonment rate appears to be substantively important by 2018 and 2019. In 2017, the Black imprisonment rate in the synthetic control was 18.99 per 100,000 population higher than in Cook County, but given the high rates of Black imprisonment this only represented a difference of 1.59%. However, in 2018, the difference between the synthetic control unit's Black imprisonment rate and Cook County's increased, with the synthetic control's rate now 116.02 per 100,000 population (10.32%) higher. The magnitude of this association grew further in 2019, with the Black imprisonment rate now 211.48 per 100,000 population (20.32%) higher in the synthetic control than in Cook County. Placebo synthetic control estimates for the other 23 counties in this sample suggest that the associations found in 2017, 2018, and 2019 were not the result of chance (p = 0.04 in 2017; p = 0.001 in 2018; and p = 0.001 in 2019).

Latinx imprisonment rates

In our last analysis, we examined the effect of Kim Fox's election on Latinx imprisonment rates. As mentioned previously, the pretreatment trends in Latinx imprisonment rates between Cook County and its synthetic control were not as similar as we would like (i.e., the fit was not great; the Latinx imprisonment rate was 1.47% different from Cook on average during the pretreatment period and 1.55% different in the year prior to treatment). Therefore, in Figure 4, we present results from bias-corrected synthetic control analyses. As with the prior figures, this figure provides information about the composition of the synthetic control unit. However, in this case, the solid line represents how different Cook County's Latinx incarceration rate is from its synthetic control, with numbers less than 0 indicating Latinx imprisonment rates are less in Cook County than its synthetic control and numbers greater than 0 indicating that Cook County's Latinx imprisonment rate is higher. With the bias-corrected synthetic control approach adjusting for bias in the pretreatment fit, we treat differences in Latinx imprisonment rates between Cook County and its synthetic control as the effect of Kim Foxx's election.

As Figure 4 demonstrates, the election of Kim Foxx did not have the same effect on the Latinx imprisonment rate as it did on White and Black imprisonment rates. While the Latinx imprisonment rate in Cook county was slightly lower in 2017 than the Latinx rate in the syntheitc control (the synthetic control had a Latinx imprisonment rate that was 1.59% higher than Cook's rate), by 2018 and 2019 the Latinx rate was higher in Cook than its synthetic control (the synthetic control's rate was 5.36% lower in 2018 and 1.35% lower in 2019). However, Placebo synthetic control estimates for the other 22 counties in this sample suggest that the associations found in 2017, 2018, and 2019 were not statistically significant (p = 0.86 in 2017; p = 0.86 in 2018; and p = 0.95 in 2019). Therefore, Kim Foxx's election appears to have not had any effect on Latinx imprisonment rates.¹

Supplemental analyses

To assess the effect of Kim Foxx's election, the control group should not consist of other counties that elected a progressive chief prosecutor. There are no other obvious progressive prosecutor jurisdictions in Illinois during the study period. However, Lake County is a potentially borderline case, one where the prosecutor elected at the end of 2012, Mike Nerheim, did not campaign as progressive prosecutor, has not been treated as such in media reports, but did engage in some progressive reforms, such as creating a conviction integrity unit and diversion opportunities for some individuals who had never been previously charged with a crime and were being charged with a nonviolent offense (Meadows, 2020). Therefore, to ensure that the results presented above were not impacted by the inclusion of Lake County, we conducted each of our prior analyses with Lake County excluded from the sample.

Since Lake County was not part of the synthetic control group for the total, Black, and Latinx analyses, the magnitude of these associations by definition could not change, but the statistical significance could. However, we found no evidence that it did, with Kim Foxx's election remaining associated with significant reductions in the total and Black imprisonment rate (p < 0.05), but not significantly associated with the Latinx imprisonment rate. Although Lake County was part of the synthetic control group for the White imprisonment rate, excluding it from the control group led to substantively similar conclusions as our main analyses. As shown in Appendix Figure A2, the magnitude of the decline in the White imprisonment rate remained quite similar. Moreover, while the p-value increased to 0.08 in 2017, the decline in the White imprisonment rate in 2018 and 2019 remained statistically significant (p < 0.05).

Discussion

In recent years, advocates have increasingly viewed electing progressive prosecutors as a key tool for criminal legal reform. Two of the most common aims of progressive prosecutors are reducing incarceration rates, as well as reversing the large racial/ethnic disparities in incarceration that currently exist (Davis, 2019). Despite the growing prominence of this approach, the evidence on its efficacy at achieving these aims remains limited (Mitchell et al., 2022). In this study, we add to this nascent literature by conducting the first quasi-experimental analysis of the implications of electing a progressive prosecutor for imprisonment rates overall, as well race/ethnic-specific imprisonment rates. Specifically, we examine the impact of the election of Kim Foxx (one of the first and most prominent progressive prosecutors) on Cook County, Illinois’ overall, Black, White, and Latinx imprisonment rate using a series of synthetic control analyses.

Overall, our findings are largely consistent with existing scholarship that suggests that prosecutors and prosecutorial policies and practices matter (Bourne, 2024; Johnson, 2018; Lynch et al., 2021; Pfaff, 2012) and that electing a progressive prosecutor is a potentially important means to achieving criminal legal reform (Davis, 2019; Mitchell et al., 2022). Specifically, we find that Kim Foxx's election was associated with a statistically significant reduction in the overall, Black, and White imprisonment rate, but not the Latinx imprisonment rate. In this way, our findings confirm what Mitchell and colleagues found in the only other empirical study that examined the impact of progressive prosecutors on imprisonment: that progressive prosecutors can be effective at reducing the use of prison. However, our results differ from Mitchell and colleagues (2022) in that our findings do not suggest that electing a progressive prosecutor is associated with a significant reduction in racial disparities in imprisonment. While Mitchell and colleagues (2022) found that Black-White disparities were smallest in jurisdictions with a progressive prosecutor, our study suggests that although both Black and White individuals benefitted from Kim Foxx's election, White individuals experienced a proportionally larger decline and Black-White disparities remain quite large at 15.33 to 1 (roughly 3.65% higher than the year before Kim Foxx took office). There are a number of possible reasons for this difference. The first is methodological: While Mitchell and colleagues (2022) relied on a cross-sectional design, we employed a quasi-experimental approach that allowed us to account for prior trends in our outcomes of interest. The second is geographical: Mitchell and colleagues (2022) examined jurisdictions in Florida, while our analyses focused on Illinois counties.

When considering the implications of our findings, it is important to recognize this study's limitations. First, while we employ a rigorous quasi-experimental approach, it is certainly possible that our findings are driven by some unobserved factor, since the election of a progressive prosecutor is not a random event. One possible threat to internal validity in this case is the bail reform that occurred in Cook County in late 2017. While we cannot rule out with certainty that bail reform is driving our findings, we believe it is quite unlikely for two reasons: (1) We focus on prison incarceration rates and bail reform would most directly impact jail incarceration rates and (2) Prior scholarship suggests that Cook County's bail reform did not significantly impact jail incarceration rates (Hinami et al., 2022), the type of incarceration rate it would be most likely to impact. Second, our findings are estimates of the effect that Kim Foxx's election had in Cook County, Illinois. It is unclear whether these findings would generalize to other jurisdictions and other progressive prosecutors. We believe it is important for future studies to use quasi-experimental approaches to analyze the impact of the election of other progressive prosecutors. Third, our analyses focus exclusively on imprisonment rates, but there are other outcomes that electing a progressive chief prosecutor could potentially impact, such as pretrial detention and charge reduction rates, which future research should also examine.

Notwithstanding these limitations, we believe our study has important implications for criminal legal reform. Our findings suggest both strengths and limitations to the current approach to progressive prosecution. For individuals and organizations engaged in reform efforts, electing a progressive prosecutor, with an orientation similar to Kim Foxx, has the potential to reduce imprisonment rates significantly and relatively quickly. This is critical since reducing the scope of carceral systems is a key aim of many individuals and groups interested in criminal legal change. A second key goal of many criminal legal change actors is to reduce racial and ethnic disparities in imprisonment. This aim is less likely to be met by electing the current group of individuals who are considered progressive prosecutors.

This is not to say that electing a progressive prosecutor does not have the potential to lead to reductions in racial and ethnic disparities in imprisonment. However, progressive prosecutors would have to embrace reforms that go beyond an exclusive focus on nonviolent offenses (Gottlieb et al., 2021). While people of color are overrepresented among those in prison for nonviolent offense charges, they are even more overrepresented among people in prison for violent offense charges (Carson, 2018; Gottlieb et al., 2024). People of color are also especially likely to be overrepresented among people with prior criminal records, which means that sentence enhancement provisions for individuals that are deemed to be “repeat offenders” are particularly detrimental to Black and Latinx individuals (Hester et al., 2018). Therefore, if progressive prosecutors want to reduce racial and ethnic disparities in imprisonment it is imperative for them to consider embracing reforms to how violence is punished and to the role that prior criminal record plays in criminal case outcomes.

Footnotes

ORCID iD

Aaron Gottlieb

Notes

Aaron Gottlieb is an associate professor at the Crown Family School of Social Work, Policy, and Practice at the University of Chicago. His research examines the structural determinants of criminal legal involvement, the consequences of criminal legal involvement, and the implications of criminal legal change.

Matt Epperson is an associate professor at the University of Chicago Crown Family School of Social Work, Policy, and Practice, where he also serves as Director of the Smart Decarceration Project. His primary areas of focus include addressing risk factors for criminal legal involvement among persons with mental illnesses, as well as advancing evidence-based approaches to effective and sustainable decarceration.

Laura Giugno is a recent graduate of the University of Chicago, where she received a BA in Political Science; Philosophy; and Law, Letters, and Society.

Hye-Min Jung is a doctoral student at the School of Social Work, Columbia University. Her research interests include racial disparity, social mobility and improving the welfare of low-income families.

Appendix

References

Abadie

(2021) Using synthetic controls: feasibility, data requirements, and methodological aspects. Journal of Economic Literature 59(2): 391–425.

Abadie

Diamond

Hainmueller

(2015) Comparative politics and the synthetic control method. American Journal of Political Science 59(2): 495–510.

Abadie

L’hour

(2021) A penalized synthetic control estimator for disaggregated data. Journal of the American Statistical Association 116(536): 1817–1834.

Ali

(2020) Kim Foxx maintains her ‘tough on crime’ message, avoiding more controversy. Medill Reports Chicago. Available at: https://news.medill.northwestern.edu/chicago/kim-foxx-maintains-touch-on-crime-message-avoiding-more-controversy/

Arora

(2018) Too Tough on Crime? The Impact of Prosecutor Politics on Incarceration. American Economic Association. Available at: https://static1.squarespace.com/static/5c8e59f6e8ba44fdeb42f85f/t/64a0e4cfe7fbef49bc4a5b7a/1688265935733/Ashna+Arora.pdf

Bartos

Kubrin

(2018) Can we downsize our prisons and jails without compromising public safety? Findings from California's prop 47. Criminology & Public Policy 17(3): 693–715.

Bartos

McCleary

Mazerolle

, et al. (2020) Controlling gun violence: Assessing the impact of Australia’s gun buyback program using a synthetic control group experiment. Prevention Science 21: 131–136.

Bazelon

(2020) Charged: The New Movement to Transform American Prosecution and End Mass Incarceration. New York, NY: Random House Trade Paperbacks.

Beckett

(2018) Mass incarceration and its discontents. Contemporary Sociology: A Journal of Reviews 47: 11–22.

10.

Bellin

(2018) Reassessing prosecutorial power through the Lens of mass incarceration. Michigan Law Review 116: 835.

11.

Bellin

(2019) The power of prosecutors. New York University Law Review 94(2): 172–212.

12.

Ben-Michael

Feller

Rothstein

(2021) The augmented synthetic control method. Journal of the American Statistical Association 116(536): 1789–1803.

13.

Bocanegra

(2019). Correctional boundaries: Examining punishment in community; a spatial analysis of probation in Chicago, Illinois . ProQuest Dissertations & Theses Global. (AAT 13901760).

14.

Bourne

(2024) The decarcerative impact of prosecutor-led charging reform. Working Paper. Berkeley Institute for Data Science.

15.

Bureau of Justice Statistics (n.d.) Prosecutors offices. Washington, DC: U.S. Department of Justice. Available at: https://www.bjs.gov/index.cfm/content/dcrp/content/data/index.cfm?ty=tp&tid=27

16.

Carson

(2018) Prisoners in 2016. Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.

17.

CDC (n.d.) Bridged-Race Population Estimates 1990–2020. Available at: https://www.cdc.gov/nchs/nvss/bridged_race.htm

18.

Clear

(2009) Imprisoning Communities: How Mass Incarceration Makes Disadvantaged Neighborhoods Worse. New York: Oxford University Press.

19.

Davis

(2019) Reimagining prosecution: a growing progressive movement. UCLA Criminal Justice Law Review 3(1): 2–27.

20.

Enns

(2014) The public’s increasing punitiveness and its influence on mass incarceration in the United States. American Journal of Political Science 58(4): 857–872.

21.

Epperson

Pettus-Davis

(2017) Smart decarceration: guiding concepts for an era of criminal justice transformation. In: Epperson

Pettus-Davis

(eds) Smart Decarceration: Achieving Criminal Justice Transformation in the 21st Century. New York, NY: Oxford University Press, 3–19.

22.

Fair

Walmsley

(2021) World Prison Population List. 13th ed. London: Institute for Criminal Policy Research.

23.

Fair and Just Prosecution (n.d.) 21 Principles for the 21^st Century Prosecutor. Available at: https://www.fairandjustprosecution.org/staging/wp-content/uploads/2018/12/FJP_21Principles_Interactive-w-destinations.pdf

24.

Ferman

Pinto

Possebom

(2020) Cherry picking with synthetic controls. Journal of Policy Analysis and Management 39(2): 510–532.

25.

Fryer

(2020) Race, reform, and progressive prosecution. The Journal of Criminal Law and Criminology 110: 769–802.

26.

Gajwani

Lesser

(2019) The hard truths of progressive prosecution and path to realizing the movement's promise. New York Law School Law Review 64: 69–94.

27.

Galiani

Quistorff

(2017) The synth runner package: utilities to automate synthetic control estimation using synth. The Stata Journal 17(4): 834–849.

28.

Gottlieb

(2017) Incarceration and relative poverty in cross-national perspective: The moderating role of female employment rates and welfare state generosity. Social Service Review 91(2): 293–318.

29.

Gottlieb

(2022) The unwarranted disparity statement: a new tool to reduce disparities in post-arrest case outcomes. UCLA Criminal Justice Law Review 6: 107–134.

30.

Gottlieb

Charles

McLeod

, et al. (2021) Were California’s decarceration efforts smart? A quasi-experimental examination of racial, ethnic, and gender disparities. Criminal Justice and Behavior 48(1): 116–134.

31.

Gottlieb

Harper

Jung

(2024) Does racial impact statement reform reduce Black–White disparities in imprisonment: Mixed methods evidence from Minnesota. Law & Policy. https://doi.org/10.1111/lapo.12252 .

32.

Gottschalk

(2015) Caught: The Prison State and the Lockdown of American Politics. Princeton, NJ: Princeton University Press.

33.

Hester

Frase

Roberts

, et al. (2018) Prior record enhancements at sentencing: unsettled justifications and unsettling consequences. Crime and Justice 47(1): 209–254.

34.

Hinami

Epperson

Lyons

, et al. (2022) Does bail reform decarcerate mental illness? Public health challenges for a large county jail system. Journal of Correctional Health Care 28(4): 267–273.

35.

Hogan

(2022) De-prosecution and death: a synthetic control analysis of the impact of de-prosecution on homicides. Criminology & Public Policy 21(3): 489–534.

36.

Illinois Department of Corrections (n.d.) Prison Population Datasets. https://idoc.illinois.gov/reportsandstatistics/prison-population-data-sets.html

37.

Johnson

(2018) In search of the missing link: examining contextual variation in federal charge bargains across U.S. Justice Quarterly 35(7): 1133–1165.

38.

Kreag

(2017) Prosecutorial analysis. Washington University Law Review 94(4): 771–826.

39.

Levin

(2021) Imagining the progressive prosecutor. Minnesota Law Review 105(3): 1415–1452.

40.

Lynch

(2023) Prosecutors as punishers: a case study of trump-era practices. Punishment & Society 25(5): 1312–1333.

41.

Lynch

Barno

Omori

(2021) Prosecutors, court communities, and policy change: The impact of internal DOJ reforms on federal prosecutorial practices. Criminology; An Interdisciplinary Journal 59(3): 480–519.

42.

McCleary

McDowall

Bartos

(2017) Design and Analysis of Time Series Experiments. Oxford: Oxford University Press.

43.

McClelland

Mucciolo

(2022) An Update on the Synthetic Control Method as a Tool to Understand State Policy. Washington, DC: Tax Policy Center.

44.

McGraw

(2021) Prosecuting with compassion, defending with power: progressive prosecutors and the case for rehabilitative justice. Penn State Journal of Law & International Affairs 9(2): 262–301.

45.

Meadows

(2020) Patch Candidate Q&A: Mike Nerheim, Lake County State's Attorney. Available at: https://patch.com/illinois/lakeforest/patch-candidate-q-mike-nerheim-lake-county-states-attorney

46.

Mitchell

Mora

Sticco

, et al. (2022) Are progressive chief prosecutors effective in reducing prison use and cumulative racial/ethnic disadvantage? Evidence from Florida. Criminology & Public Policy 21(3): 535–565.

47.

Morrison

(2020) Progressive Prosecutors Scored Big Wins in 2020 Elections, Boosting a Nationwide Trend. The Conversation. Available at: https://theconversation.com/progressive-prosecutors-scored-big-wins-in-2020-elections-boosting-a-nationwide-trend-149322

48.

Neal

Rick

(2023) The role of policy in prison growth and decline. Working Paper No. 2023-150. Available at: https://bpb-usw2.wpmucdn.com/voices.uchicago.edu/dist/6/3381/files/2023/12/SSRN-id4649647.pdf

49.

Nellis

(2021) The Color of Justice: Racial and Ethnic Disparity in State Prisons. Washington, DC: The Sentencing Project. Available at: https://www.sentencingproject.org/reports/the-color-of-justice-racial-and-ethnic-disparity-in-state-prisons-the-sentencing-project/

50.

Neyfakh

(2015) Why did it take more than a year to charge the officer who shot Laquan McDonald? Slate. Available at: https://slate.com/news-and-politics/2015/11/laquan-mcdonald-kim-foxx-on-why-anita-alvarez-mishandled-the-jason-van-dyke-case.html

51.

Nichols

(2021) The police union failed miserably in its attempt to beat Philadelphia DA Larry Krasner. The Nation. Available at: https://www.thenation.com/article/politics/larry-krasner-carlos-vega-win/

52.

Petersen

Mitchell

Yan

(2024) Do progressive prosecutors increase crime? A quasi-experimental analysis of crime rates in the 100 largest counties, 2000–2020. Criminology & Public Policy 23(2): 459–490.

53.

Pfaff

(2012) The micro and macro causes of prison growth. Georgia State University Law Review 28: 1237–1271.

54.

Pfaff

(2017) Locked in: The True Causes of Mass Incarceration and How to Achieve Real Reform. New York: Basic Books.

55.

Pickerell

(2020) Critical race theory and power: the case for progressive prosecution. Harvard Blackletter Law Journal 36: 73–89.

56.

Reclaim Chicago, The People’s Lobby, and Chicago Appleseed Fund for Justice (n.d.) In pursuit of justice for all: An evaluation of Kim Foxx’s first year in office. Available at: https://www.chicagoappleseed.org/wp-content/uploads/2017/12/In-Pursuit-of-Justice-For-All_Kim-Foxx-Report_For-Web.pdf

57.

Reclaim Chicago, The People’s Lobby, Chicago Council of Lawyers, and Chicago Appleseed Center for Fair Courts (2021) Decarcerating Cook County: Use of diversion programs for low-level, “non-violent” felonies increased substantially under State’s Attorney Kim Foxx compared to Anita Alvarez. Available at: https://www.chicagoappleseed.org/wp-content/uploads/2021/03/22021-03-CCSAO-Diversion-Report-FINAL.pdf

58.

Rice

(2018) Kim Foxx just released six years of data – most prosecutors’ offices remain black boxes. The Appeal. Available at: https://theappeal.org/kim-foxx-just-released-six-years-of-data-most-prosecutors-offices-remain-black-boxes-238a37ee45f0/

59.

Richardson

Kutateladze

(2021) Tempering expectations: a qualitative study of prosecutorial reform. Journal of Research in Crime and Delinquency 58(1): 41–73.

60.

Sabol

Johnson

(2022) Justice System Disparities: Black-White National Imprisonment Trends, 2000 to 2020. Washington, DC: Council on Criminal Justice. Available at: https://counciloncj.foleon.com/reports/racial-disparities/national-trends

61.

Schmadeke

(2016) Foxx raises threshold for felony shoplifting. Chicago Tribune. Available at: https://digitaledition.chicagotribune.com/tribune/article_popover.aspx?guid=9ac9f3b1-0978-4ba4-a81b-bca34e82a2fa

62.

Schultz

(2023) Philly’s progressive prosecutor, facing impeachment trial, has authority on transit crimes diverted. AP. Available at: https://apnews.com/article/philadelphia-district-attorney-septa-prosecutor-79be3e6e7f3b7d0c8050b65703be776e

63.

Sklansky

(2017) The progressive prosecutor’s handbook. UC Davis Law Review 50: 25–42.

64.

Starr

Rehavi M

(2013) Mandatory sentencing and racial disparity: assessing the role of prosecutors and the effects of booker. The Yale Law Journal 123(2): 2–80.

65.

StataCorp (2021) Stata statistical software: Release 17.

66.

Stuntz

(2006) The political constitution of criminal justice. Harvard Law Review 119: 780–851.

67.

Wakefield

Uggen

(2010) Incarceration and stratification. Annual Review of Sociology 36(1): 387–406.

68.

Wildeman

(2016) Incarceration and population health in wealthy democracies. Criminology; An Interdisciplinary Journal 54(2): 360–382.

69.

Wiltshire

(2022) allsynth: (Stacked) Synthetic control bias-correction utilities for Stata. Working paper, University of California–Davis. https://justinwiltshire.com/s/allsynth_Wiltshire.pdf .