Advanced Search
Search Terms:
Blog Posts
Data Sets
People: Profiles
Press Releases
Climate Change/Energy
Fiscal/Governance Reform
Health & Human Services
K-12 Education
Political Landscape
PPIC Higher Education Center
PPIC Statewide Survey
PPIC Water Policy Center
Any Date
Past month
Past 6 months
Past 12 months
Custom range

Clear All

Pretrial Detention and Jail Capacity in California


California’s persistently overcrowded jails are facing additional challenges now that public safety realignment has shifted many lower-level offenders from state prisons to county supervision. Jail capacity challenges are prompting a reconsideration of California’s heavy reliance on holding unsentenced defendants in jail pending trial—known as pretrial detention. The legal rationale for pretrial detention is to ensure court appearances and preserve public safety. But California’s high rates of pretrial detention have not been associated with lower rates of failure to appear or lower levels of felony rearrests. This report concludes that pretrial services programs—if properly implemented and embraced by the courts, probation, and the jails—could address jail overcrowding and improve the efficiency, equitability, and transparency of pretrial release decision making.

Many of California’s Jail Inmates Are Unsentenced

Overcrowding has long been a problem for many California jails. For more than a decade, a number of counties have operated facilities under court orders that mandate the release of inmates at specified population thresholds (Lawrence 2014; Welsh 1995). At last count, 39 jail facilities in 19 counties were operating under such population caps (Martin and Lofstrom 2014). These caps have compelled the monthly release of thousands of inmates who would otherwise have remained in custody (Figure 1).

Figure 1. Jail overcrowding has led to the release of thousands of inmates

Figure 1 Sample

SOURCE: Board of State and Community Corrections, Jail Profile Survey Jan 2004–September 2014.

NOTE: Unsentenced inmate releases include only those inmates who would not be released if there were sufficient jail capacity.

The potential of pretrial release programs to relieve jail population pressures becomes apparent when we examine jail populations more closely. The most recently available data show that as of September of 2014, roughly 50,000—or 62 percent—of beds were filled with inmates awaiting trial or sentencing (Figure 2).1 Moreover, the need for jail space for sentenced inmates has been growing since 2012, after the state implemented corrections realignment.

Figure 2. Most jail inmates are not serving sentences

Figure 1 Sample

SOURCE: Board of State and Community Corrections, Jail Profile Survey Jan 2004–September 2014.

NOTE: Unsentenced inmates are awaiting either trial or sentencing.

Given that unsentenced inmates make up such a large share of the jail population, the appeal of programs to improve the management of pretrial detention space is readily apparent.2 Before we consider how county pretrial services programs—most of which have been implemented since realignment—could improve pretrial release decision making, we need to look at the way these decisions are made with and without pretrial services programs.

Pretrial Release Decision Making

In California, the sequence of events following an arrest depends on the severity of the arrest offense and on local policy and practice. Generally, for minor offenses, law enforcement has the discretion to cite and release an arrestee in lieu of taking that person into custody. By signing the citation, the arrestee is promising to appear in court at a specified time and place. Law enforcement usually transports arrestees suspected of committing more serious offenses to the local jail for booking—but local policies, practices, and court orders grant jail officials discretion to book and release some lower-level arrestees on signed promises to appear in court.3

If a suspect is not released after booking, he or she may be eligible for release by posting bail.4 Alternatively, at a defendant’s first court appearance, generally within 48 hours of arrest, the judge may grant a recognizance release, which is a no-bail release requiring a signed promise to appear, or similarly a conditional release, which is a recognizance release with the imposition of added conditions such as electronic monitoring.5 The judge may also adjust the bail amount.6 If the judge does not grant a recognizance release or conditional release, the defendant is returned to custody and retains the right to post bail. Nearly all who secure financial releases in California do so through bail bonds. In the rest of the nation, deposit bonds and, to a lesser extent, full cash bonds are more common.7

Assessing and Managing Risk with Pretrial Services Programs

Pretrial services programs have two major roles: to collect and present information that will assist the court in making pretrial release decisions and to provide monitoring and supervisory services for released defendants.8 The use of these programs and the program models employed vary widely across the state; some counties—such as Santa Cruz, Santa Clara, and San Francisco—have longstanding pretrial services programs, while other counties have implemented or reestablished programs in the wake of realignment (see Technical Appendix A). Counties vary in the extent to which their programs align with national standards (ABA 2007).

Pretrial services programs have two major roles: to collect and present information that will assist the court in making pretrial release decisions and to provide monitoring and supervisory services for released defendants.
For example, some counties have yet to implement the supervision component of these programs. Counties also differ in the information they provide to the courts, but in most instances that information includes the results of objective risk assessments. Finally, counties differ in the extent to which courts have embraced the principles and goals of pretrial services programs.

Assessing Risk

Pretrial services programs begin by conducting a risk assessment prior to a defendant’s first court appearance.9 In general, statistically or actuarially based tools are used to assess the probability of pretrial misconduct—namely, failure to appear in court and/or new criminal activity. These tools generate risk scores based on a set of factors that have the strongest empirically demonstrated relationship to pretrial misconduct. For example, the Ohio Risk Assessment Systems Pretrial Assessment Tool includes risk factors such as age at first arrest, number of prior failure-to-appear warrants and jail incarcerations, employment at the time of arrest, residential stability, illegal drug use in the previous six months, and severe drug use problems (LaTessa et al. 2009). The consensus among researchers is that these tools produce risk scores that are more consistent and more accurate than unaided professional judgment (Mossman 2010; Andrews et al. 2006; Ægisdottir et al. 2006).

Mitigating Risk

A core principle of pretrial services programs is that defendants should be released under the least restrictive conditions that will reasonably ensure court appearance and public safety, and that supervision and compliance monitoring should be responsive to the defendant’s assessed risk (ABA 2007). Pretrial services agencies may recommend and provide a range of services and supervision options to mitigate pretrial risks, including court date reminders to reduce failures to appear in court; referrals to drug, alcohol, or mental health services; phone or in-person check-ins; or electronic monitoring services (Pretrial Justice Institute 2009; Clark and Henry 2003).10 Regardless of the recommendation made by the pretrial service program, the court is ultimately responsible for imposing conditions of release.

Assessing and Mitigating Risk without the Support of Pretrial Services Programs

In the absence of pretrial services programs, judges in California assess defendant risk based on their professional judgement. As prescribed by law, when making pretrial release decisions, judges consider the safety of the public, the victim, and the victim’s family, as well as the seriousness of the offense, the defendant’s criminal record, and the probability that the defendant will appear in court.11 However, the law provides no mechanism for systematically weighing these factors, nor any requirement of consistency in their use among judges.12 In the absence of pretrial supervision options, judges also have limited ability to assign pretrial release conditions that mitigate risk. This may have contributed to high pretrial detention rates prior to the recent implementation of many pretrial services programs.

California’s High Rates of Pretrial Detention

From 2000 to 2009 (the latest comprehensive data available for felony cases), California’s large urban counties relied on pretrial detention to a much greater extent than did large urban counties elsewhere in the United States (Figure 3).13 The share of felony defendants for whom bail is denied in California is similar to the share in the rest of the United States.14 Part of the difference in detention rates may be attributed to California’s higher bail amounts. The median bail amount in California ($50,000) is more than five times the median amount in the rest of the nation (less than $10,000).15 Research has demonstrated that pretrial release rates generally decline as bail amounts increase (Cohen and Reaves 2007).

Figure 3. California has relied more heavily than the rest of the United States on pretrial detention

Figure 1 Sample

SOURCE: U.S. Department of Justice, Bureau of Justice Statistics. State Court Processing Statistics 1990–2009: Felony Defendants in Large Urban Counties.

NOTE: Unweighted results from 2000–2009 are presented. For comparison of weighted and unweighted results see Technical Appendix B. "Rest of the US” refers to the full national data set with California counties excluded.

Also contributing to California’s relatively high pretrial detention is the prevalence of defendants on probation or parole, with prior convictions, and with prior failures to appear. However, even after differences in demographics and criminal history are accounted for, California still detains defendants at a significantly higher rate than the rest of the United States.16 Despite these relatively high rates of pretrial detention, California’s rates of failure to appear and nonviolent felony rearrest are no lower than rates in the rest of the United States (Table 1). In fact, California has higher rates of multiple failures to appear in court, especially for those who secure nonfinancial releases.17

Table 1. California has had higher rates of failures to appear and higher rearrests for non-violent felonies

Figure 1 Sample

SOURCE: US Department of Justice, Bureau of Justice Statistics. State Court Processing Statistics, 1990–2009: Felony Defendants in Large Urban Counties.

NOTE: Unweighted results from 2000–2009 are presented. For comparison of weighted and unweighted results see Technical Appendix B. "Rest of the US” refers to the full national data set with California counties excluded.

Pretrial misconduct is particularly high among defendants charged with felony drug offenses (Table 2). Now that Proposition 47 has downgraded some non-serious, non-violent drug possession and property crimes to misdemeanors, California’s rates of felony rearrests and failure to appear in court on felony charges may decrease. However, if law enforcement simply cites and releases suspects for Proposition 47 offenses, pretrial misconduct such as rearrests or recitations are likely to follow, as are high rates of failures to appear in these misdemeanor cases.

Table 2. Drug offenders in California have had higher rates of pretrial misconduct

Figure 1 Sample

SOURCE: U.S. Department of Justice, Bureau of Justice Statistics. State Court Processing Statistics 1990–2009: Felony Defendants in Large Urban Counties.

NOTE: Presented here are unweighted results from 2000–2009. Pairwise tests for multiple FTAs show that differences are significant (at 0.01 level) for each comparison of drug offense type with other offense types. Pairwise tests for felony rearrests for a violent offense (contingent on rearrest) show significant (at 0.01 level) differences for each comparison of violent offense arrest with other offense types. See Technical Appendix B for a more detailed list of offenses included in each of the offense categories presented.

Given that the legal rationale for pretrial detention is to ensure court appearances and preserve public safety, the data presented here indicate that California may not be getting a good return on the high levels of pretrial detention it has maintained. California’s pretrial practices are associated with lower rates of rearrests for violent felonies, but this result may have been achieved at the cost of detaining many defendants who might have safely been released under some form of pretrial supervision. Moreover, as critics of the bail system have long argued, releasing defendants based on their ability to post bail is both inequitable and unnecessarily risky: defendants with financial resources can purchase release even if there is a high risk that they will engage in pretrial misconduct, while low-risk defendants who are poor may be needlessly held in jail.

Improving Pretrial Decision Making

California’s recently passed Proposition 47, which downgraded some non-serious, non-violent drug possession and property crimes to misdemeanors, has eased some of the state’s jail overcrowding issues in the short term.18 But counties will probably need a range of options, including pretrial services programs, to overcome long-term jail capacity challenges.19 Because these programs rely on detention only for those defendants who pose the highest risk of pretrial misconduct, using other supervision strategies for those posing a lower risk, they offer a promising approach to handling the state’s high levels of pretrial detainees.

Pretrial services programs also have the potential to make pretrial decision making more transparent, efficient, and equitable. Without these programs, the chance of a defendant being released from jail while awaiting trial depends on a number of factors, including jail capacity, the financial means to secure bail, and the discretion of multiple actors in the criminal justice system. Because these factors are not related to individual risk in a manner that lends itself to measurement and improvement, relying on them to handle pretrial populations may not be the most effective management strategy.

The approach taken by pretrial services programs focuses pretrial decision making squarely on risk of pretrial misconduct and lends itself to measurement and improvement. If properly implemented, risk assessment tools yield systematic, objective, and consistent measures of risk. Because the underlying algorithms of these programs are open to examination, they also represent a more transparent alternative to California’s traditional approach to pretrial decision making. If the principles and objectives of these programs can be agreed upon system wide, they offer counties an opportunity to better manage pretrial populations while also reinforcing the integrity of their local justice systems.


  1. Author’s calculations: Board of State and Community Corrections, Jail Profile Survey Second Quarter Calendar Year 2014.
  2. For a look at alternatives to incarceration for sentenced inmates, see Martin and Grattet (2015).
  3. California penal code §853.6 allows law enforcement to release most of those arrested on misdemeanor charges either in the field or after booking with a signed notice to appear in court. Those arrested for lower-level felonies may be booked and "cited out” in facilities that are under court orders related to jail overcrowding. Policies with regard to releases due to overcrowding vary from county to county.
  4. For a defendant’s right to bail and exceptions see: penal code § 1271 and California Constitution article 1 §§ 12, 28 (f)(3). For discharge from custody on bail see: §§ 1268 ,1269, 1269b. For local court responsibility for bail schedule and basis for bail amounts based on seriousness of charge see: PC 1269b(e). For discussion of California’s bail system and variation in bail schedules see: Sonya Tafoya, Assessing the Impact of Bail on California’s Jail Population (Public Policy Institute of California, June 2013).
  5. PC 1270 (a), and California Constitution article 1 § 28 (f)(3). California penal code § 1319(b) prohibits "release on own recognizance” (ROR) if the offense charged is violent and if the defendant has failed to appear on a prior felony. California penal code § 1319(c) requires a statement of reasons for denial or granting of ROR for violent felonies. California penal code § 1318 requires that the defendant sign a promise to appear, that includes a promise to obey all conditions, to not leave the state without permission of the court, a waiver of extradition, and a statement of understanding of the consequences of violating conditions. California Penal code § 1318(a)(2) authorizes the imposition of "reasonable conditions” on a defendants own recognizance release. California penal code § 1318 allows for the use of investigative staff for the purpose of recommending whether the defendant should be released on his or her own recognizance. For further discussion see: §55.53 of Bail and Own-Recognizance Release (Administrative Office of the Courts 2013).
  6. PC 1269(c) governs requests for bail deviation.
  7. Deposit bonds are issued to a defendant who deposits a share of the full amount, usually 10 percent, with the court. If the defendant appears for all hearings, the deposit, minus a small administrative fee, is refunded. If the defendant fails to appear, he or she is liable for the full bond amount. Full cash bonds require the defendant to deposit the full amount with the court. If the defendant appears in court, the full amount is refunded. If the defendant fails to appear, he or she is liable for the full bond amount.
  8. For more detail, see American Bar Association (ABA) Standard 10-1.10 and National Association of Pretrial Service Agencies (NAPSA) Standard 1.3 (a).
  9. Generally, this information is collected though an interview; however, a risk assessment tool that uses only standardized information from criminal history records has been piloted in multiple jurisdictions across the nation and is currently being piloted in Santa Cruz and San Francisco (VanNostrand and Lowenkamp 2013).
  10. Court date notifications or reminders have proved effective in reducing failures to appear, according to a range of studies, but more research is needed to establish effective practices and interventions for mitigating pretrial misconduct (VanNostrand et al. 2011).
  11. California Constitution article 1 § 28 (f)(3), and article 1, §28(b)(3) and penal code § 1275(a).
  12. Research on judicial decision making in New York demonstrated that in addition to prosecutors’ bail request and charge severity, the individual judge in the case had a significant impact on the decision to release or detain and the bail amount set (Phillips 2012).
  13. Data contained in the State Court Processing Statistics (SCPS) track the movement of felony defendants through the criminal courts. The data are drawn from approximately 40 of the nation’s 75 largest urban counties in even numbered years from 1990–2006 and in 2009 and include information on arrest offenses, demographics, criminal history, pretrial processing, adjudication, and sentencing of felony defendants. See Technical Appendix B for detailed information on the SCPS. The SCPS data pre-date the implementation dates of most pretrial services programs in California.
  14. Author’s calculations of unweighted data show that California denies bail to 5 percent of defendants, compared with 6 percent in the rest of the United States.
  15. Author’s calculation with unweighted data from SCPS 2000–2009.
  16. See Technical Appendix B for regression results.
  17. Recognizance and conditional release are two of the three types of nonfinancial release listed in the State Court Processing Statistics; the third type is an unsecured bond. Unsecured bonds require no payment on the part of defendants who make all required court appearances. The defendant becomes liable for the bond amount only if he or she fails to appear in court. In California, the vast majority of nonfinancial releases are recognizance releases.
  18. See "Prop. 47's Effect on Jail Time, Drug Rehabilitation Is Mixed So Far,” Los Angeles Times, February 21, 2015, and "Prop. 47 Report Finds Fewer Drug Arrests, Less Crowding in Jails,” February 25, 2015. The Board of State and Community Corrections has not released post-election data on capacity releases contained in the Jail Profile Survey.
  19. See: Martin and Lofstrom (2014) for discussion of projected jail construction needs.
Back to top


Administrative Office of the Courts. 2013. Bail and Own-Recognizance Release, California Judges Benchguides, Benchguide 55.

Ægisdottir, Stefania., et al. 2006. "The meta-analysis of Clinical Judgement Project: Fifty-six Years of Accumulated Research on Clinical versus Statistical Prediction.” Counseling Psychologist 34: 341–82.

American Bar Association. 2007. ABA Standards for Criminal Justice – Pretrial Release, 3rd edition. Washington, DC.

Andrews, D. A., James Bonta, and J. Stephen Wormith. 2006. "The Recent Past and Near Future of Risk and or Need Assessment.” Crime & Delinquency 52 (1).

Clark, John, and D. Alan Henry. 2003. Pretrial Services Programming at the Start of the 21st Century A Survey of Pretrial Services Programs, U.S. Department of Justice Office of Justice Programs Bureau of Justice Assistance.

Crime and Justice Institute. 2014. Monterey County, California, Pretrial Justice System Assessment, Crime and Justice Institute, June 16.

Cohen, T. H., and B. A. Reaves. 2007. Pretrial Release of Felony Defendants in State Courts. Washington, DC: U.S. Department of Justice, Office of Justice Programs.

Danford, Kristy Pierce, and Meghan Guevara. 2013. Creating An Effective Pretrial Program a Toolkit for Practitioner, Crime and Justice Institute, July 2013.

Latessa, Edward, Paula Smith, Richard Lemke, Matthew Makarios, Christopher Lowenkamp. 2009. Creation and Validation of the Ohio Risk Assessment System. Final Report. University of Cincinnati School of Criminal Justice Center for Criminal Justice Research.

Lawrence, Sarah. 2014. Court-Ordered Population Caps in California County Jails, Stanford Criminal Justice Center, Stanford Law School.

Martin, Brandon, and Magnus Lofstrom. 2014. Key Factors in California’s Jail Construction Needs. Public Policy Institute of California.

Martin, Brandon, and Ryken Grattet. 2015. Alternatives to Incarceration in California. Public Policy Institute of California.

Mossman, Douglas. 2010. "Understanding Risk Assessment Instruments.” In The American Psychiatric Publishing Textbook of Forensic Psychiatry, ed. Simon and Gold. Washington: American Psychiatric Publishing, Inc.

National Association of Pretrial Services Agencies. 2004. Standards on Pretrial Release, 3rd edition. October 2004.

Partnership for Community Excellence. 2012. Pretrial Detention and Community Supervision Best Practices and Resources for California Counties.

Phillips, Mary T. 2012. A Decade of Bail Research in New York City. New York City Criminal Justice Agency, Inc.

Pretrial Justice Institute. 2009. Pretrial Justice in America: A Survey of County Pretrial Release Policies, Practices and Outcomes. Pretrial Justice Institute.

U.S. Department of Justice, Bureau of Justice Statistics. State Court Processing Statistics, 1990-2009; Felony Defendants in Large Urban Counties [Computer file]. Conducted by Pretrial Justice Institute [producer] ICPSR2038-v5. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor].

State of California, Board of State and Community Corrections, Jail Profile Survey Second Quarter Calendar Year 2014 Survey Results.

VanNostrand, Marie, and Christopher T. Lowenkamp. 2013. Assessing Pretrial Risk without a Defendant Interview. Laura and John Arnold Foundation.

VanNostrand, Marie, Kenneth J. Rose, and Kimberly Weibrecht. 2011. "The State of the Science of Pretrial Release Recommendations and Supervision.” Pretrial Justice Institute.

Welsh, Wayne. 1995. Counties in Court, Jail Overcrowding and Court Ordered Reform, Temple University Press, 1st edition.


The author gratefully acknowledges the thoughtful reviews of Judge Richard Couzens (Ret.), Thomas Cohen, Jacob Jackson, and Ryken Grattet. Any errors in this work are my own.



Fiscal/Governance Reform