Tilting the Scales Charge and Juror Data Dictionary

Tilting the Scales Charge and Juror Data Dictionary

In order to examine how Louisiana’s unusual law allowing non-unanimous verdicts was impacting the criminal justice system in the state, The Advocate undertook a year-long investigation with the goal of compiling as comprehensive a dataset as possible of trials, defendants and juries.

The analysis shows that at every step, from the day jury summonses go out to the final proclamation from the foreman, race is a crucial part of how the scales can be tilted against defendants in Louisiana. The Advocate’s on-going series can be found at theadvocate.com/tiltingthescales.

Data on criminal charges and juror selection in Louisiana was collected by reporters at The Advocate, with the intention of developing a dataset that covered as many trials as possible. The data largely covers 2011 – 2016 though it also includes some records outside that range.

Nearly all the data was collected from court minutes, other public records or media reports on trials as almost no parishes in the state do not provide machine-readable data on trials or jurors.

Data collection focused on the busiest court districts in the state, eventually collecting substantial numbers of records from nine out of the top 10 parishes by caseload. The tenth, Lafayette Parish, refused to comply with requests for information from The Advocate.

Additional data was gathered on trials in another 35 parishes, though in many cases those parishes could not provide detailed records on those trials.

The charges dataset includes information on 3,028 trials involving 3,120 defendants and 5,394 individual charges. Of those, 2,931 cases occurred during the 2011 – 2016 time frame, out of a total of 3,906 cases in those years reported to the Louisiana Supreme Court.

Given the focus of the investigation, special attention was paid to gathering information on whether verdicts were unanimous or non-unanimous.While it is typical for attorneys call for that information after a verdict is rendered, it is not always requested and the results are not always recorded in court records.

Of the cases in the dataset, it was possible to determine whether verdicts were unanimous on 993 convictions out of the 2,027 cases that ended with at least one guilty verdict from a jury. Those cases cover half of the state’s 64 parishes, though they are heavily weighted toward the large parishes of Orleans, Jefferson, St. Tammany, East Baton Rouge and Caddo. Collectively, those parishes are responsible for about 68 percent of the convictions in the state and roughly 69 percent of the data on jury unanimity in the dataset.

Once basic information about the trials was entered into the dataset, efforts were made to gather information on jury selection in as many of those cases as possible. The records kept about the jury selection process vary from parish to parish, and in some cases from judge to judge, which limited the ability to get a full accounting of the jury selection process

The jurors dataset includes information on 41,303 potential jurors who were called to a venire, when attorneys from both sides determine if they will sit on a jury. Information was recorded on whether the potential juror was picked to serve on the jury, was eliminated from service by an attorney on one side or the other or if they had some other outcome.

Information about juror’s race and sex was taken from court records where available and gathered from voter registration records or other databases when it was not.

That process yielded information on 939 trials with 10,995 jurors, a dataset that significantly overlaps the trials in which the unanimity of verdicts is known.

The data covers parishes that represent all geographic regions, demographic mixes and urbanization patterns found in the state.

The process of collecting data on jurors and potential jurors focused on 12-member juries, as opposed to the six-member panels that can weigh in on less serious offenses but can only convict with a unanimous vote.

In a small number of cases, full information was not available about the jurors who served on a case or the people who were questioned during a venire. In other cases, the process was halted because of a plea deal or a mistrial during jury selection. Those cases are included in this dataset for the sake of completeness, though it is recommended they be excluded for analysis.

A final round of data collection involved seeking information on the racial breakdown of the juror votes when verdicts were not unanimous. That information was only available for 109 of cases involving 157 verdicts. Almost half of those cases were in East Baton Rouge, where the Clerk of Court’s Office assisted in tallying the votes.

General description of the data:

The data consists of two csv files, one with details on defendants, charges and the outcome of those charges and the other with details on jurors in each case. The two sets of data can be joined on the trial_id fields.

If you use this data, please credit The Advocate.

If you have any questions about these datasets, please email Jeff Adelson at .

Charge data (found in TiltingTheScalesCharges.csv)

Field / Type / Description
charge_id / Integer / Identifier unique to each charge
jurisdiction_jdc / String / Judicial District Court number
jurisdiction_parish / String / Parish in which the trial occurred
trial_id / String / Identifier unique to each trial, shared by all charges in a case and all jurors in a case.
Note:
Charges against multiple defendants tried together will have the same trial_id.
jurisdiction_section / String / JDC section, where known
def_name / String / Defendant’s name
def_dob / Date / Defendant’s date of birth
def_race / String / Defendant’s race
charge_original / String / The criminal charge the defendant faced at the beginning of trial. Generally speaking, these are listed by reference to the Louisiana Criminal Code.
charge_verdict / String / The criminal charge the defendant was found guilty of, if different than charge_original
verdict / String / Description of verdict, including whether it was guilty, not guilty, a mistrial or a hung jury. Also indicates whether a guilty verdict was for a lesser charge than the original charge and whether the verdict was the result of a plea deal.
verdict_date / Date / Date verdict was rendered
verdict_unanimity / String / Was the verdict unanimous?
Null = Unknown
sentence_years / Integer / Total sentence for all charges that result in a guilty verdict against a defendant in a specific case.
8888 = Life without parole
9999 = Death penalty
deliberation_minutes / Integer / Number of minutes the jury deliberated
habitual_offender / String / Was the defendant sentenced under habitual offender guidelines?
Null = Unknown
jurors_total / Integer / Total number of jurors on the jury, where known
jurors_guilty / Integer / Total number of jurors who supported a guilty verdict, where known
jurors_dissent / Integer / Total number of jurors who voted against a guilty verdict
jurors_guilty_white / Integer / Total number of white jurors supporting a guilty verdict
jurors_guilty_black / Integer / Total number of black jurors supporting a guilty verdict
jurors_guilty_other / Integer / Total number of jurors whose race is known but who are neither white nor black supporting a guilty verdict
jurors_guilty_unknown / Integer / Total number of jurors whose race is not known supporting a guilty verdict
jurors_dissent_white / Integer / Total number of white jurors who voted against a guilty verdict
jurors_dissent_black / Integer / Total number of black jurors who voted against a guilty verdict
jurors_dissent_other / Integer / Total number of jurors whose race is known but who are neither white nor black who voted against a guilty verdict
jurors_dissent_unknown / Integer / Total number of jurors whose race is not known who voted against a guilty verdict

Juror data (found in TiltingTheScalesJurors.csv).

Note: Personally identifying information about the jurors has been removed to protect the jurors’ privacy.

Field / Type / Description
juror_id / Integer / Identifier unique to each juror
jurisdiction_parish / String / Parish in which the trial occurred
trial_id / String / Identifier unique to each trial, shared by all charges in a case and all jurors in a case.
Note:
Charges against multiple defendants tried together will have the same trial_id.
defendant_name / String / Defendant’s name.
Note:
-Defendants’ names may differ from names in charges data due to differences in collection methodology
trial_date / Date / Approximate start date of trial
juror_race / String / Race of juror. Races have been standardized to the six categories used in voter registration (Asian, Black, Hispanic, White and other). Jurors whose race could not be determined are listed as unknown.
juror_sex / String / Sex of juror, where known.
Juror_outcome / String / What happened to the potential juror:
ALTERNATE: Potential juror was selected as an alternate.
-Note: Some case records do not distinguish between alternates and other jurors. In those trials, alternates will be included as jurors.
DEFENSE CAUSE: Eliminated by the defense for cause.
DEFENSE PEREMPTORY: Eliminated by the defense with a peremptory strike.
EXCUSED: Excused from service by the court.
JOINT CAUSE: Eliminated for cause by both the defense and prosecution.
JOINT PEREMPTORY: Eliminated by peremptory strikes from both the defense and prosecution.
JUROR: Potential juror served on a jury. See note under “ALTERNATE.”
NOT QUESTIONED: Potential juror was not questioned during venire.
STATE CAUSE: Eliminated by prosecutors for cause.
STATE PEREMPTORY: Eliminated by prosecutors with a peremptory strike
UNKNOWN: Case record does not indicate what happened to this potential juror.