I’ve just been reading a 2001 study by David Mustard, of the University of Georgia, called “Racial, Ethnic and Gender Disparities in Sentencing: Evidence from the US Federal Courts.” ((The Journal of Law and Economics, vol. 44, no. 1, pages 285-314. Pdf link.)) Mustard’s study appears better-designed than other sentencing studies I’ve read. His sample is large and comprehensive: he essentially includes every federal sentence handed down for three consecutive years (1991 through 1993) in his analysis. Rather than focusing only on sex or on race, he simultaneously controls for the effects of race, sex, U.S. citizenship, and class on federal sentencing. (Legally, none of those four factors are supposed to have an effect on what sentence a judge hands down.)
The results aren’t pretty. Especially for drug crimes and for bank robberies, being white is a big advantage if you’re being sentenced for a federal crime:
Bank robbery and drug trafficking exhibit the largest black-white differentials. Blacks receive 9.4 and 10.5 months longer than whites in bank robbery and drug trafficking, respectively. The percentage difference is greatest for those convicted of drug trafficking, where blacks are assigned sentences 13.7 percent longer than whites. The aggregate Hispanic-white difference is driven primarily by those convicted of drug trafficking and firearm possession/trafficking, the only two crimes with significant Hispanic coefficients. For these two crimes, Hispanics receive 6.1 and 3.7 additional months compared to whites, or 8.0 percent and 7.0 percent longer in percentage terms.
Note that Mustard’s analysis only compared felons who were convicted for the same crime. So the above sentencing disparities do not include the infamous disparities caused by the much harsher sentences given for crack cocaine possession (usually a Black crime) than for powder cocaine possession (usually a white crime).
Being a woman is an even larger advantage for bank robbers:
The female-male difference is statistically significant for all six categories, the largest of which is for bank robbery, where females receive 21.6 months less than males.
Although the bank robbery differential was largest, women received a break on sentencing compared to men across the board. ((Mustard’s report didn’t include a discussion of the death penalty, but it appears that women are less likely to receive the death penalty than similarly-situated men. See, for example, Victor L. Streib (2006), “Rare and Inconsistant: The Death Penalty For Women,” Fordham Urban Law Journal, Vol 33 pp 609.))
I was particularly surprised that controlling for dependents didn’t significantly alter the male/female difference – so the sentencing disparity is apparently not being caused by judges taking mercy on single mother defendants.
Class made a significant difference, but mainly for the very poor. That is, people who earned less than $5,000 a year get fewer breaks in sentencing than people who earn more than $5,000 a year; but there doesn’t appear to be much difference in the sentences given those who earn $10,000 a year and those who earn $50,000. An exception was sentencing for fraud: “Those with incomes greater than $50,000 receive significantly shorten sentences for fraud.”
Finally, being a U.S. citizen leads to lighter sentencing across the board.
Having no high school diploma resulted in an additional sentence of 1.2 months. Income had a significant impact on the sentence length. Offenders with incomes of less than $5,000 were sentenced most harshly. This group received sentences 6.2 months longer than people who had incomes between $25,000 and $35,000. Those with U.S. citizenship receive lower sentences by about 1.7 months, perhaps because they take advantage of their greater knowledge about the court systems and legal representation. Age is positively related to the sentence length. [...]
The income and education results could be generated if people with higher levels of education and income use their resources to obtain more favorable sentences. However, if offenders utilize education and income to reduce their sentences, the impact is limited. The marginal productivity of income in hiring legal resources diminishes quickly after income hits a minimum threshold, because individuals with the highest incomes do not receive reductions in sentence length.
According to Mustard’s analysis, most of the sentencing disparities are caused by judges departing from the official sentencing guidelines; when judges decide to take mercy on a felon and offer a very light sentence – or to not sentence the felon to prison at all – they are significantly more likely to do so if the felon is not poor, is white, is female, and is a U.S. citizen.
Feminist readers are likely to take particular note of the harsher sentences given to men, compared to women convicted of the same crimes. Mustard suggests that this may be caused by sexist paternalism among judges; women are seen less as full adults, and as being less capable of being responsible for their own actions, and as a result judges depart from sentencing guidelines to give women lighter sentences. Although I can’t know if that’s true or not, it certainly seems plausible to me, and also compatible with feminist analysis of how women are treated and viewed by society.
Another study, by Max Schanzenbach of the Northwestern University School of Law, ((Max M. Schanzenbach, “Racial and Gender Disparities in Prison Sentences: The Effect of District-Level Judicial Demographics” (April 2004). American Law & Economics Association Annual Meetings. American Law & Economics Association 14th Annual Meeting. Working Paper 4. Pdf link.)) looked at sex disparities in sentencing according to the sex of the judge. He found that, for serious crimes, female judges did not give harsher sentences to men, but male judges did:
The greater the percentage of female judges on a district’s bench, the smaller the gender disparity. These results are hard to square with the suggestion that unobserved accomplice status or blameworthiness is behind the gender disparity. At the very least, male and female judges view the dangerousness, accomplice status, or blameworthiness of female offenders differently.
The female offender/percent female judge effects [...] were not evident at all in the category of less serious crimes. (There was some evidence in the case of less serious crimes that more Democratic districts treated men and women alike when granting downward departures.) However, paternalistic views about the dangerousness or blameworthiness of female offenders may well be most evident in the case of serious crimes.
Schanzenbach also found that racial and ethnic disparities were only slightly decreased, or not decreased at all, in districts with more black and hispanic judges. However, he argues that this finding does not prove a lack of racial and ethnic bias in sentencing, only that if such bias exists, it’s not dependent on the race or ethnicity of the judge.