The Anatomy of Discretion: An Analysis of Prosecutorial Decision-making for Cases Processed by Offices in One Northern County and One Southern County, 2007-2010 (ICPSR 32542)

Principal Investigator(s): Frederick, Bruce, Vera Institute of Justice; Stemen, Don, Loyola University of Chicago

Summary:

Prosecuting attorneys enjoy broad discretion in making decisions that influence criminal case outcomes. This study examines the impact of legal, quasi-legal, and extra-legal factors on case outcomes throughout the prosecutorial process. It then examines how prosecutors weigh these factors in their decision making and explores the formal and informal mechanisms that constrain or regulate prosecutors' decision-making.

The study examines case screening decisions, charging decisions, plea offers, sentence recommendations, and dismissals in two moderately large county prosecutors' offices. It includes statistical analyses of actual case outcomes, responses to a standardized set of hypothetical cases, and responses to a survey of prosecutors' opinions and priorities, as well as qualitative analyses of two waves of individual interviews and focus groups. It addresses the following questions:

  1. How did prosecutors define and apply the concepts of justice and fairness?
  2. What factors were associated with prosecutorial outcomes at each stage?
  3. How did prosecutors interpret and weigh different case-specific factors in making decisions at each stage?
  4. How did contextual factors constrain or regulate prosecutorial decision making?
  5. How consistent were prosecutors' decisions across similar cases? What case-level and contextual factors influenced the degree of consistency?

Access Notes

  • One or more files in this data collection have special restrictions ; consult the restrictions note to learn more. You can apply online for access to the restricted-use data. A login is required to apply.

    Access to these data is restricted. Users interested in obtaining these data must complete a Restricted Data Use Agreement, specify the reason for the request, and obtain IRB approval or notice of exemption for their research.

    Any public-use data files in this collection are available for access by the general public. Access does not require affiliation with an ICPSR member institution.

Dataset(s)

DS1:  Northern County Screening Data
Documentation:
Download:
No downloadable data files available.
DS2:  Northern County ADA Data
Documentation:
Download:
No downloadable data files available.
DS3:  Northern County Dismissals Data
Documentation:
Download:
No downloadable data files available.
DS4:  Northern County Focus Group Qualitative Data
Download:
No downloadable data files available.
DS5:  Southern County Person and Property Case-Level Data
Documentation:
Download:
No downloadable data files available.
DS6:  Southern County Person and Property Charge-Level Data
Documentation:
Download:
No downloadable data files available.
DS7:  Southern County Person and Property Victim-Level Data
Documentation:
Download:
No downloadable data files available.
DS8:  Southern County Drug Case-Level Data
Documentation:
Download:
No downloadable data files available.
DS9:  Southern County Drug Charge-Level Data
Documentation:
Download:
No downloadable data files available.
DS10:  Southern County Modified Charge Code Table Data
Documentation:
Download:
No downloadable data files available.
DS11:  Southern County Focus Group Qualitative Data
Download:
No downloadable data files available.
DS12:  Factorial Survey ADA-Level Input to HLM Data
Documentation:
Download:
No downloadable data files available.
DS13:  Factorial Survey Case-Level Input to HLM Data
Documentation:
Download:
No downloadable data files available.
DS14:  Factorial Survey Case-Level Data
Documentation:
Download:
No downloadable data files available.
DS15:  Factorial Survey ADA-Level Data
Documentation:
Download:
No downloadable data files available.
DS16:  Hypothetical Defendant Profiles for Factorial Survey Data
Documentation:
Download:
No downloadable data files available.
DS17:  Combined Sites General Survey Data
Documentation:
Download:
No downloadable data files available.

Study Description

Citation

Frederick, Bruce, and Don Stemen. The Anatomy of Discretion: An Analysis of Prosecutorial Decision-making for Cases Processed by Offices in One Northern County and One Southern County, 2007-2010 . ICPSR32542-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2015-12-23. https://doi.org/10.3886/ICPSR32542.v1

Persistent URL: https://doi.org/10.3886/ICPSR32542.v1

Export Citation:

  • RIS (generic format for RefWorks, EndNote, etc.)
  • EndNote XML (EndNote X4.0.1 or higher)

Funding

This study was funded by:

  • United States Department of Justice. Office of Justice Programs. National Institute of Justice (2009-IJ-CX-0040)

Scope of Study

Subject Terms:    case dismissal, case management, case processing, court cases, disposition (legal), district attorneys, guilty pleas, judicial process, plea negotiations, pretrial procedures, prosecuting attorneys, prosecution, trials

Smallest Geographic Unit:    county

Geographic Coverage:    United States

Time Period:   

  • 2009-01--2011-06 (Northern County Administrative Data, screened between January 2009 and June 2011)
  • 2007-01--2009-07 (Southern County Administrative Data; Drug Cases May 2007 - July 2009, felony person and property cases January 2007 - June 2007)
  • 2011 (Hypothetical Case Vignettes: Fall of 2011)
  • 2010 (Opinion survey: Fall of 2010)
  • 2010-10--2011-03 (Qualitative Data: Southern County: October 2010 and December 2010; Northern County: November 2010 and March 2011.)

Date of Collection:   

  • 2011 (Hypothetical Case Vignettes: Fall of 2011)
  • 2010 (Opinion survey: Fall of 2010)
  • 2010-10--2011-03 (Qualitative Data: Southern County: October 2010 and December 2010; Northern County: November 2010 and March 2011.)

Unit of Observation:    Individuals (Datasets 2, 7, 12, 15, 16, 17), Charges (Datasets 1, 3, 6, 9, 10), Complaint-defendant (case) (Datasets 5, 8, 13, 14)

Universe:    All Assistant District Attorneys (ADAs) in Northern County and Southern County (Parts 2, 4, 11, 12-17). All arrest charges in Northern County screened between January 2009 and June 2011 (Parts 1, 3). All felony cases forwarded from the largest urban police agency in Southern County and screened between January 1, 2007 and June 30, 2007(Parts 5, 6, 7). All felony drug cases screened by Southern County from May 1, 2007 through July 31, 2009 (Parts 8, 9).

Data Type(s):    administrative records data, observational data, survey data

Data Collection Notes:

Users should be aware that the two participating District Attorney's offices are identified only by the pseudonyms "Northern County" and "Southern County".

The qualitative data are available through restricted access procedures in two zip files. The file qda32542-0004_REST.zip (Part 4, Northern County Focus Group Qualitative Data) includes notes from five focus groups conducted with ADAs and team captains, and two interviews with the District Attorney (DA). The file qda32542-0011_REST.zip (Part 11, Southern County Focus Group Qualitative Data includes notes from 4 focus groups conducted with ADAs and new team leaders, and 3 interviews with the DA and Deputy DA.

Users should be aware that the feedback received from prosecutors in the focus groups (Parts 4 and 11) was recorded as field notes. Since interviews/focus groups were not recorded, these notes cannot produce exact quotes in all instances.

Users should be aware that there are some discrepancies in case counts between the information in the final report for the study, and the data available in this collection. For Part 5, the final report indicates that there was a final sample of 658 unique cases from the property team. However, only 654 cases are selected using the FILT4PROPERTYCOHORT filter found in the data. The 658 count was obtained using a different filter, "FILT4MOST_NODRGORSEX" before the principal investigators discovered four homicide cases that didn't belong in the dataset. The filters "FILT4PROPERTYCOHORT" and "FILT4PERSONCOHORT" excluded those four cases.

Methodology

Study Purpose:   

This study examines case screening decisions, charging decisions, plea offers, sentence recommendations, and dismissals in two moderately large county prosecutors' offices. It includes statistical analyses of actual case outcomes, responses to a standardized set of hypothetical cases, and responses to a survey of prosecutors' opinions and priorities, as well as qualitative analyses of two waves of individual interviews and focus groups. It addresses the following questions:

  • How did prosecutors define and apply the concepts of justice and fairness?
  • What factors were associated with prosecutorial outcomes at each stage?
  • How did prosecutors interpret and weigh different case-specific factors in making decisions at each stage?
  • How did contextual factors constrain or regulate prosecutorial decision making?
  • How consistent were prosecutors' decisions across similar cases? What case-level and contextual factors influenced the degree of consistency?

Study Design:   

This study used a multi-method approach, relying on both quantitative and qualitative methods to examine prosecutorial decision making in Northern and Southern Counties. These included: two waves of interviews and focus groups (Datasets 4 and 11), a general survey (Datasets 12, 15, and 17), a factorial survey (Datasets 12-16), and an analyses of administrative data (Datasets 1-3, 5-10).

To examine actual case outcomes and the factors associated with those outcomes, the study relied on analyses of administrative data derived from the case management systems in each office and a review of case files in Southern County. The study further explored case outcomes using a factorial survey containing a series of hypothetical cases in which prosecutors were asked to make and justify screening, charging, and plea offer decisions. The analyses of actual and hypothetical case outcomes were complemented by an attitudinal survey of prosecutors and series of interviews/focus groups with prosecutors. Prosecutors in both jurisdictions were invited to respond to a structured survey, designed to elicit general attitudes about definitions of individual and organizational success; the influence of relationships among prosecutors, police, defense attorneys, and judges; resource and policy constraints; principles that guide screening decisions and plea offers; the general goals and functions of the criminal justice system; and internal training and oversight. Two waves of interviews with the District Attorney (DA) and Deputy District Attorneys (Deputies) in each site and two waves of focus group sessions with line prosecutors and unit managers also were conducted in each of the research sites.

Sample:   

The study relies on data from two county prosecutors' offices - Northern County and Southern County.

The analyses of case outcomes in Northern County relied on individual-level administrative data collected in the case management system of the district attorney's office. The initial sample (Northern County Screening Data, Dataset 1) included 111,704 unique arrest charges. The Northern County ADA Data (Dataset 2) contains one record for each Assistant District Attorney (ADA) who screened any of the cases in Dataset 1, for a total of 126 cases. The Northern County Dismissals Data (Dataset 3) contains one record for each charge accepted for prosecution in Northern County, for a total of 30,535 cases. The Northern County Focus Groups Qualitative Data (Dataset 4) included five prosecutors responsible for different types of cases (general crimes, domestic violence, drugs, weapons) and seven unit managers.

In Southern county, the person and property unit data (Datasets 5, 6, and 7) include all closed felony cases initially forwarded from the largest urban police agency in Southern County and screened between January 1, 2007 and June 30, 2007. Homicide cases and cases in which the top charge was for kidnapping or any sex offense were excluded. The Southern County Person and Property Case-Level Data (Dataset 5) includes 1,319 cases, defined as a complaint-defendant, the Southern County Person and Property Charge-Level Data (Dataset 6) includes 2,905 charges, and the Southern County Person and Property Victim-Level Data (Dataset 7) includes 1,739 victim charges. The drug unit data (Datasets 8 and 9) include all closed felony drug cases screened by the office from May 1, 2007 through July 31, 2009, including 5,423 complaint-defendants (Dataset 8, Southern County Drug Case-Level Data) and 13,931 charges (Dataset 9, Southern County Drug Charge-Level Data). The Southern County Modified Charge Code Table Data (Dataset 10) has data for the 1,520 specific, statutorily defined offenses in Southern State. Southern County Focus Groups (Dataset 11) included six prosecutors with less than one year experience, eight prosecutors with one to ten years' experience, and seven unit managers.

For the factorial survey (Datasets 12-15), in Southern County, participation was solicited from 67 prosecutors who had previously responded to the general survey. Sixty-two prosecutors completed the survey (Datasets 12 and 15). Each respondent answered questions about 10 different case vignettes, yielding up to 620 observations for each question (Datasets 13 and 14). Due to a low response rate in Northern County (only 18 responses were received initially and a concerted follow-up effort yielded only two additional responses), only the factorial survey data from Southern County were analyzed and included in this collection. The Hypothetical Defendant Profiles for Factorial Survey Data (Dataset 16) includes 780 cases, each of which corresponds to a unique combination of packet, vignette, criminal history profile, and race.

For the Combined Sites General Survey Data (Dataset 17), a total of 155 prosecutors responded. This includes 74 respondents from Southern County and 81 prosecutors from Northern County.

Time Method:    Cross-sectional , Time Series: Continuous

Mode of Data Collection:    record abstracts, face-to-face interview, mail questionnaire

Data Source:

District Attorney's automated case management system in Northern County (Datasets 1, 2, 3)

Paper case files in Southern County (Dataset 5, 6, 7)

District Attorney's drug case data base in Southern County (Datasets 8, 9, 10)

In-person interviews with the District Attorney (DA) and Deputy District Attorneys (Deputies) in each site and focus group interviews with Assistant District Attorneys and unit managers (Datasets 4 and 11).

Description of Variables:   

Northern County Screening Data (Dataset 1) has 24 variables, including demographic characteristics of defendant and victim, charge details, and whether the charge was rejected or accepted for prosecution. The variable named "MOSTSERIOUSARREST" distinguishes between the most serious arrest charge and other charges.

Northern County ADA Data (Dataset 2) has 15 variables, including information on Assistant District Attorneys (ADA) gender, role (supervisor/manager or not; specialized or general unit), length of tenure in the office, and the distribution of case types screened by the ADA.

Northern County Dismissals Data (Dataset 3) has 24 variables, including information similar to that in Dataset 1, plus indicators of whether the charge was subsequently dismissed and whether the entire case was dismissed.

Northern County Focus Group Qualitative Data (Dataset 4) and Southern County Focus Group Qualitative Data (Part 11) have qualitative data from two waves of focus groups and interviews. The first wave focused primarily on contextual conditions and circumstances that influence decision making (e.g., guiding philosophies; policies; relationships with other system actors and colleagues; resource constraints). The second wave of interviews/focus groups focused primarily on case-specific factors that influence decision making (e.g., strength of evidence, seriousness of the instant offense, defendant's criminal history, and special aggravating/mitigating circumstances).

Southern County Person and Property Case-Level Data (Dataset 5) has 567 variables, including case-level information on defendant demographic characteristics and criminal history; case outcomes for screening decisions, lower court decisions, formal plea offers and sentence recommendations, and upper court decisions; physical evidence listed on the police report; summaries of victim characteristics and willingness to prosecute; and detailed charge information summarized to the case level for each stage in processing.

Southern County Person and Property Charge-Level Data (Dataset 6) has 35 variables, including crime type; statutory classification; charge disposition at screening, lower court, plea offer, and upper court; reasons for charge dismissal at screening, in lower court, plea offer, and upper court; and a measure of "incarceration exposure," calculated as the upper end of the presumptive range of the minimum sentence in months if convicted of the charge.

Southern County Person and Property Victim-Level Data (Dataset 7) has 31 variables, including victim's demographic characteristics, victim injury, victim-defendant relationships, and whether the victim was willing to prosecute.

Southern County Drug Case-Level Data (Dataset 8) has 549 variables, including case-level information on defendant demographic characteristics and criminal history; case-level outcomes for screening decisions, initial charging decisions, decisions in lower court, grand jury preparation, indictment, formal plea offers, decisions in felony administrative court, dispositions at trial, detailed summaries of charge-level information at most of these stages, and reasons for dismissals at each stage.

Southern County Drug Charge-Level Data (Dataset 9) has 217 variables, including charge-level outcomes for screening decisions, initial charging decisions, indictment, formal plea offers, decisions in felony administrative court, dispositions at trial, reasons for dismissals at each stage, and a measure of "incarceration exposure," calculated as the upper end of the presumptive range of the minimum sentence in months if convicted of the charge.

Southern County Modified Charge Code Table Data (Dataset 10) has 25 variables, including charge details such as statutory category (felony, misdemeanor, traffic offense, or infraction), detailed statutory class (A, B, C, etc.), an associated Uniform Crime Reporting (UCR) code, charge ranking values, citation of the applicable statute, and a literal description of the offense. In also includes the calculated incarceration exposure measure.

The Factorial Survey ADA-Level Input to HLM Data (Dataset 12) and Factorial Survey ADA-Level Data (Dataset 15) have 26 variables and 150 variables respectively, including information about the prosecutor, and responses to items on the general survey. Dataset 15 includes variable about age, race, gender and marital status of the prosecutor, his or her team assignment, and the length of experience in their current team and the DA's office. In addition, the file contains each prosecutor's responses to the items on the general survey, as well as three short scales constructed from combinations of the general survey items--orientation toward severity of outcome, orientation toward fairness, and orientation toward consistency. Dataset 12 has a subset of the variables in Dataset 15, with pre-calculated cross-product terms for testing selected case-level interactions.

Factorial Survey Case-Level Input to HLM Data (Dataset 13) and Factorial Survey Case-Level Data (Dataset 14) have 27 and 64 variables respectively, taken from data yielded by the factorial survey. Dataset 14 has all of the raw data yielded by the factorial survey, including researcher's design classification of offense seriousness and strength of evidence; prosecutors' rating of strength of evidence and defendant criminal history; whether the prosecutor would reject or accept the case for prosecution; reasons for rejection; details of the top filing charge; details of the top plea offer charge; top charge incarceration exposure; and sentence recommendation. Dataset 13 has a subset of the variables in Dataset 14, with pre-calculated cross-product terms for testing selected case-level interactions.

Hypothetical Defendant Profiles for Factorial Survey Data (Dataset 16) has 68 variables, representing the defendant profile data base that was used to generate the race and criminal history in the Factorial Survey.

Combined Sites General Survey Data (Dataset 17) has 329 variables, including responses to all of the items on the survey and responses to a separate background questionnaire providing prosecutors' demographic characteristics and information about the length of their experience in the office and in their current unit. The general survey incorporated a total of seventy-six items, organized in eight substantive categories: 1) factors that define professional success for individual prosecutors (fifteen items); 2) factors that define success for the district attorney's office (nineteen items); 3) the influence of relationships among prosecutors, police, defense attorneys, and judges on decision making (ten items); 4) resource and policy constraints (thirteen items); 5) principles that guide screening decisions (six items); 6) principles that guide the development of plea offers (eight items); 7) general goals and functions of the criminal justice system (nine items); and 8) training and oversight (sixteen items). Responses to all but two of the items were ratings on five-point scales that reflected either perceived frequency of occurrence, degree of agreement, or perceived importance, as appropriate. (Responses to the remaining two items involved categorical choices rather than quantitative ratings.) The raw item responses were also converted to a series of z-scores relating an ADA's responses to each item within a category to that ADA's average responses to the items in that category.

Response Rates:   

For the factorial survey (Datasets 12-15), in Southern County, participation was solicited from 67 prosecutors who had previously responded to the general survey. A total of 62 prosecutors completed the survey (Datasets 12 and 15) for a response rate of 93 percent.

For the Combined Sites General Survey Data (Dataset 17), in Southern County, survey responses were received from 74 respondents from a pool of 78 prosecutors (95 percent response rate). In Northern County, responses were received from 81 prosecutors from a pool of 135 prosecutors (60 percent response rate).

Presence of Common Scales:    Several Likert-type scales were used (Datasets 12, 15, and 17)

Extent of Processing:   ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major statistical software formats as well as standard codebooks to accompany the data. In addition to these procedures, ICPSR performed the following processing steps for this data collection:

  • Created variable labels and/or value labels.
  • Standardized missing values.
  • Checked for undocumented or out-of-range codes.

Version(s)

Original ICPSR Release:   2015-12-23

Version History:

  • 2016-04-21 Update to the User Guide to correct for missing Publications and NIJ Data Resources Program sections.

Related Publications

Variables

Utilities

Metadata Exports

If you're looking for collection-level metadata rather than an individual metadata record, please visit our Metadata Records page.

Download Statistics