Examining Policy Impacts on Racial Disparities in Federal Sentencing Across Stages and Groups and over Time, [United States], 1998-2021 (ICPSR 38647)

Version Date: Apr 25, 2024

Cite this study |

Principal Investigator(s):
Mari McGilton, Urban Institute

https://doi.org/10.3886/ICPSR38647.v1

Version V1

Slide tabs to view more

Summary

In this secondary analysis, the research team used publicly available federal sentencing data from the United States Sentencing Commission (USSC) to measure racial disparities for multiple race groups and stages of sentencing across time (fiscal years 1999-2021). They sought to answer the following research questions:

  1. Do racial disparities vary across 3 stages of federal sentencing and over time? If so, how?
  2. During which years do the measured racial disparities have a statistically significant decrease?
  3. Which policies likely impacted these decreases the most? What are the commonalities between them?

To answer the research questions, the research team measured racial disparities between matched cases across three stages of federal sentencing, represented by two elements each; identified at which points in time the disparities changed significantly using time series plots and structured break analyses; and used this information to systematically review federal policies to identify which might have contributed to significant decreases in racial disparities.

This collection contains 1 analytic dataset (n = 1,281,732) containing 27 key variables for all fiscal years and the code/syntax used to complete the secondary analysis:

  • 5 files to compile and clean the original data and produce matched datasets (3 R, 1 SAS, 1 Stata)
  • 6 files to analyze sentences by race (all R)
  • 4 files to analyze sentences by federal sentencing guideline (all R)
  • 11 files to analyze sentences by circuit court (all R)

Please refer to the Data Sources metadata field and accompanying documentation for details on obtaining the original data.

Citation

McGilton, Mari. Examining Policy Impacts on Racial Disparities in Federal Sentencing Across Stages and Groups and over Time, [United States], 1998-2021. Inter-university Consortium for Political and Social Research [distributor], 2024-04-25. https://doi.org/10.3886/ICPSR38647.v1

Export Citation:

  • RIS (generic format for RefWorks, EndNote, etc.)
  • EndNote

Funding

United States Department of Justice. Office of Justice Programs. National Institute of Justice (15PNIJ-21-GG-02817-RESS)

Subject Terms

Geographic Coverage

Smallest Geographic Unit

Federal court district

Distributor(s)

Inter-university Consortium for Political and Social Research
Hide

Time Period(s)

1998-10-01 -- 2021-09-30 (Federal fiscal years 1999 through 2021)

Date of Collection

2022-04-01 -- 2022-04-30

Data Collection Notes

  1. Please visit the United States Sentencing Commission (USSC) website for more information about U.S. federal sentencing data and to download data and documentation for fiscal year 2002 onward.
  2. This collection is related to the Monitoring of Federal Criminal Sentences Series. Please visit the ICPSR series website for more information about the series and access to data and documentation for individual studies.
Hide

Study Purpose

The purpose of the study was to measure racial disparities across stages in the United States federal sentencing process, detecting when statistically significant changes in disparities occurred and identifying which policies impacted significant decreases the most.

Study Design

Publicly available individual-level federal sentencing data for fiscal years (FY) 1999, 2000, and 2001 were downloaded from the National Archive of Criminal Justice Data (NACJD) through ICPSR. Data for FY2002 through 2021 were downloaded from the United States Sentencing Commission (USSC) website. Cases where the information on sentencing documents matched were kept. Cases that had missing/illogical values or where the maximum possible sentence was death were excluded. The final dataset created for analysis had a case count of 1,281,732.

RStudio's Matchit package was used to perform exact matching on the data to limit as many legal and extralegal factors that contribute to federal sentencing decisions as possible. The research team used regressions to measure the size of disparities between matched racial groups for each of the three federal sentencing stages (charge, computation, and sentence). The R package strucchange was used to detect structural breaks (breakpoints) in the trend of differences for each stage across fiscal years. Subgroup analyses were also conducted to remeasure disparities between racial groups, offense type/federal sentencing guidelines, and circuit court. Two researchers with subject matter expertise in federal sentencing policy and racial disparities reviewed legislation, Supreme Court decisions, sentencing guidelines amendments, and U.S. attorney general directives enacted during (and up to 2 years prior to) breakpoint dates identified in the subgroup analyses and in the Black/non-Black racial group analysis. Each policy was rated on the likely influence it had on the observed decrease in disparities on a scale of 1 to 5 (low to high influence).

Time Method

Time Series

Universe

Individual cases sentenced federally and published by the United States Sentencing Commission (USSC) and/or National Archive of Criminal Justice Data (NACJD).

Unit(s) of Observation

Event/Process (Federally Sentenced Case)

Data Source

Data Type(s)

Mode of Data Collection

Description of Variables

Racial groups were defined based on United States Sentencing Commission (USSC) data collection: white, Black, American Indian and Alaska Native (AIAN), and Asian and Pacific Islander (API). An "other" racial category was not examined separately due to the variety of identities falling into that group. Hispanic origin was used as a separate covariate.

Dependent measures included 6 elements across 3 stages of federal sentencing: charge stage (mandatory-minimum eligibility, statutory maximum eligible sentence), computation stage (final offense level, total criminal history points), and sentence stage (total sentence length, number of months sentenced below the maximum of the guideline range). Final offense level and sentence length were not limited. The maximum number of months in the guideline range was limited to 470 (a life sentence is defined as 470 months by the USSC).

Matching covariates were age group, sex, primary sentencing guideline, district of sentencing, fiscal year, education level, Hispanic origin, U.S. citizenship status, and whether the case was resolved via trial or guilty plea. For analyses of total sentence length and number of months below the guideline-range maximum, final offense level and criminal history category were added as matching covariates.

Hide

Original Release Date

2024-04-25

Hide

Notes

  • The 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.