Independent and Joint Effects of Race, Gender, and Age on Federal Sentencing Outcomes: Analysis of the 2018-2019 United States Federal Sentencing Data

The current study explored the independent and joint effects of race, gender, and age on federal sentencing outcomes to identify if non-legal factors-individually or in combination- influence sentencing decisions at the federal level. Logistic regression analysis was conducted on the 2018-2019 Monitoring of Federal Criminal Sentences data (n = 76,119) collected by the United States Sentencing Commission to explore the independent and joint effects of race, gender, and age on sentencing outcomes. Due to the nuances of judicial discretionary decision-making, the non-legal factors that can influence sentencing outcomes must be studied both independently and jointly. Controlling for offense type, prior criminal history, method of conviction, U.S. citizenship, and educational attainment levels, results show that the individual effects of being male, younger, and non-White made it more likely to be sentenced to prison over probation. Additionally, results show that the joint effects of race, age, and gender resulted in young non-White males and older non-White males to have higher odds of being sentenced to prison over probation. Continuing education for legal actors, such as judges, who are primarily responsible for the final sentence, and prosecutors, who choose which charges to pursue, about these disparities is an important step in helping to eliminate the disparities. The dataset lacks information about additional individual characteristics such as socioeconomic status, marital status, and having a dependent. It also lacks information about case characteristics such as victim characteristics, presence of a weapon, and presence or severity of the victim's injury. All of these nonlegal factors could potentially influence judicial discretionary decision-making and sentencing outcomes. Thus, future research should further include a more extensive set of variables in the analysis of sentencing outcomes. Additionally, replication is needed using new sentencing data to track trends over time.