« "Breaking Down Barriers: Experiments into Policies That Might Incentivize Employers to Hire Ex-Offenders" | Main | Detailing how AG Sessions seeks to block sentencing reforms in White House criminal justice reform push »

January 17, 2018

Lies, damn lies and fascinating statistics in the US Sentencing Commission FY 2017 sentencing data

I just noticed that the US Sentencing Commission last week released its latest standard quarterly data report, and this one is extra exciting because it contains preliminary data on all cases sentenced during fiscal year 2017.  Critically, FY17 runs October 1, 2016 through September 30, 2017, so a good chunk of the data reflect a period in which Attorney General Loretta Lynch was still in charge of the Justice Department.  Still, a majority of the data reflects sentencings after Attorney General Jeff Sessions took over, and the final third of FY 2017 had all sentencings taking place after AG Sessions issued his May 2017 charging and sentencing memorandum directing federal prosecutors to more regularly seek within-guideline sentences.

I provide all this backstory largely as a prelude to highlighting how similar the USSC FY17 data look to FY16 data. I also thought it interesting to compare some of these data to FY13 and FY09, the last two Prez election year USSC data sets. (I am drawing all these data from Table 19, then Table 6 of these USSC data reports.)

USSC FY        Total Sentences (mean in month)     Drug Trafficking Sentences (mean in month)     Immigration Sentences (mean in month)

2009                81,347 (47 months)                                  23,931 (78 months)                                         25,924 (17 months)

2013                80,035 (45 months)                                  22,354 (72 months)                                         24,972 (16 months)

2016                67,740 (44 months)                                  19,231 (66 months)                                         20,052 (13 months)

2017                66,409 (45 months)                                  18,980 (70 months)                                         20,333 (12 months)

One can mine a lot more data from the FY 2017 report to tell a lot more stories about how, at least so far, formal and informal changes by AG Sessions have not yet made a dramatic impact on federal sentencing statistics.  Indeed, one might be heartened by the fact that fewer federal cases were sentenced in FY 2017 than in the last 15 years, and I think fewer federal drug trafficking sentences were imposed in FY17 than in nearly any other year in the past two decades (though the uptick in average sentence is interesting and may prompt a future post). 

Of course, these data may start looking very different in FY 2018 and beyond as new US Attorneys appointed by Prez Trump take over and their new cases make it all the way to sentencing. Still, I think it notable and interesting that the first run of federal sentencing data of the Trump Era shows a continued decline in overall sentences imposed and in drug trafficking sentences imposed.

January 17, 2018 at 11:13 AM | Permalink


It takes a long time to get from a plea to sentencing in federal court (and obviously a lot longer from initial charge to sentencing). The Sessions memo on charging and sentencing policy came out in May 2017, so you would not expect to see much effect on sentences imposed before September 30, 2017. I would not read anything into these statistics on the impact of the current administration. Look again in a year.

Posted by: Mark Weintraub | Jan 17, 2018 1:15:33 PM

I largely agree on your lag-time point, Mark, but I think some other factors are also in play (e.g., an Obama-heavy judiciary).

Posted by: Doug B. | Jan 17, 2018 4:26:46 PM

The FBI Index covers common law crimes, in the jurisdictions of the states. Internet hacking and identity theft are federal crimes. There are 15 million a year. The majority of sentences are given for drug dealing and illegal entry. That leaves around 30,000 criminal sentences for every other federal crime, including tax evasions, fraud/abuse, 100's others.


But assume all 30,000 represent sentences for internet crimes. That means there is a 1 in 500 chance of being inconvenienced by the criminal justice system of the federal government. The average take is $5000, compared to $4000 for the average bank robbery.

Posted by: David Behar | Jan 17, 2018 9:04:54 PM

This report from 2016 also has some great nuggets. https://www.ussc.gov/sites/default/files/pdf/research-and-publications/federal-sentencing-statistics/guideline-application-frequencies/2016/Use_of_SOC_Offender_Based.pdf

In the entire year in 2016 there were the following number of sentences in the entire U.S.A.:

Some federal criminal prosecutions are threatened incessantly but very rarely result in a conviction and sentence.

82 criminal copyright/trademark cases
40 perjury cases
1 failure to appear by a material witness
1 possession of a dangerous weapon while boarding an aircraft

Then there is the thicket of child pornography cases. Only murder and kidnapping have average sentences longer than child porn and that are far fewer of both kinds of cases in the federal courts combined than there are child pornography cases. Sexual assault and racketeering, for example, have shorter sentences. So does possession of a firearm by a felon. Only 273 child porn cases involved sexual acts involving a minor by the offender About 80% of child porn trafficking cases (the predominant charge) did not involve an exchange of money or a thing of value.

Posted by: ohwilleke | Jan 18, 2018 2:55:00 AM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB