AEJG built and maintains this Equity Dashboard for your use. To start, explore our data sources and glossary, or select a use case to learn more about the Equity Dashboard types.
Governments, nonprofit organizations, and businesses rely on the U.S. population information provided decennially through the Census. Over time, the Census Bureau has changed the way it classifies race and ethnicity. Historically, the changes have been influenced by social, political and economic factors including emancipation, immigration and civil rights. More information can be found at www.census.gov.
CFC data is available for the years 1999-2021. Outlier cases in the CFC data have been filtered out of AEJG’s sentencing views (550 cases with sentences > 9998 months). Washington’s legislature and administrative agencies use CFC data to inform their understanding of caseloads and budgets. The data, available at WA Caseload Forecast Council, includes all adult felony cases in Washington Superior Courts. The CFC obtains their data, including demographics, from the forms prepared at each felony sentencing hearing.
This data is used despite data limitations that can impact proportionality analysis and mask disparities.
According to information provided in September, 2022, the CFC supplements its data with that maintained by the Washington State Patrol and Administrative Office of the Courts. But if there are conflicts in these data sets, the CFC defaults to its own data.
The following illustrates how data limitations can skew our understanding of disproportionality. Better race/ethnicity data would help provide answers to these questions. Better data would make disproportionality, or lack thereof, more transparent.
AEJG supports improved data keeping by the CFC and others in the criminal legal system. AEJG is also working to add more data sets to the Equity Dashboard to address limitations inherent in any one dataset. Currently, AEJG is incorporating approximately 50 years of data from the Washington State Administrative Office of the Courts (AOC).
Established in 1957 by state lawmakers, the Washington State Administrative Office of the Courts (AOC) maintains information and records from court case management systems across Washington’s non-unified courts, including the Judicial Information System (JIS), the Superior Court Management and Information System (SCOMIS), the Appellate Court System (ACORDS), and Odyssey. AOC uses the data, in part, to develop operational budgets, evaluate judicial system functionality, maintain the computer system that serves the entire court system, provide continuing judicial education, compile statewide court statistics, and provide information to the judicial community, other branches of the government, and the public. The AOC tracks data from points at all stages of felony cases filed in any of Washington’s 39 superior courts.
AEJG does not expect that adding AOC data will remedy all of the limitations found in the CFC data. That will require improved data keeping by actors throughout the criminal legal system.
The population of a particular demographic group within the criminal legal system should closely mirror the representation of that same group in the general population. By overlaying Caseload Forecast Council data with Washington State census data, the Equity Dashboard displays any disproportionality that may exist.
To use this dashboard, select race, year, counties, and/or gender to see the percentage of that group in the conviction data compared to the census/population data.
If you select all counties, the tool will automatically display the 5 counties that have the highest population of the selected race to display in the Ratio Trend view.
A disproportionality ratio of 1 is represented by the blue dashed line on the Ratio Trend view.
Washington State Census data does not identify Latinx as a category, so Latinx data is included in “other.”
Gender options are binary as those are the only categories collected in the dataset.
In Washington State, felony sentencing is governed by the Sentence Reform Act of 1981 (SRA) sentencing grid. The grid accounts for the seriousness of an offense and different levels of criminal history, providing the judge with a sentencing length range. Sentencing disparity is a circumstance where similar cases are not treated similarly or different cases are not treated differently.
We have converted the sentencing data into an average sentence range percentage. If the sentencing grid requires a sentencing range of 9 to 12 months for a specific scenario:
To use this dashboard, select county, judge, race, gender, offense, and/or date range to see the average sentencing ranges. If you would like to explore the actual numbers of cases, you can scroll across the Cases & Average Sentence Range Percentage field.
Average sentence range percentages that are 50% or higher are highlighted in red.
Instances where a judge has sentenced 20 or fewer cases are highlighted in red.
Note: Users may occasionally see a sentence that is more than 100% of the range (enhancement(s) or an exceptional sentence). There also may be situations where a sentence is less than 0% of the range. These are both because they are exceptional sentences below or above the range.
This dashboard allows you to review the sentence range percentage information by race, age groups, seriousness level of the crime, and offender score. To use this dashboard, select county, judge, gender, offense, and/or date range.
Average sentence range percentages that are 50% or higher are highlighted in red.
Instances where a judge has sentenced 20 or fewer cases are highlighted in red.
This dashboard allows for further sorting by offense date, verdict type, age range, sentence length, seriousness level, and offender score. It also allows you to see the individual case numbers, sentence length, offense descriptions, and more.
This dashboard allows to you to apply a statistical chi-square test to the conviction proportionality data. A chi-square test will indicate if the comparison is statistically significant. Statistical significance allows us to understand if the difference observed between proportion of convictions and proportion of the population is likely or unlikely due to random noise in the data.
