« Alabama struggling with enduring challenges as tough-on-crime history creates "box of dynamite" | Main | SCOTUS finally to grapple with how states are applying Atkins »

March 2, 2014

Has anyone formally calculated exactly how very few federal sentences are found unreasonable?

The question in the title of this post is prompted by two noteworthy (or perhaps not really noteworthy) circuit opinions from last week in which two way-above-guideline sentences were affirmed as reasonable by the panels of the First and Seventh Circuits in US v. Santiago-Rivera, No. 13-1228 (1st Cir. Feb. 28, 2014) (available here); US v. Castaldi, No. 10-3406 (7th Cir. Feb. 24, 2014) (available here).  In both cases, a defendant appealed as unreasonable the imposition of a prison sentence roughly a decade longer than the (already lengthy) prison term suggested by the applicable guidelines.  In both cases, the panel found the way-above-guideline sentence procedurally and substantively reasonable (though one Seventh Circuit judge dissented in Castaldi).

At one level, these two rulings highlight how increased district court discretion to sentence outside the guidelines will often be used enhance sentences involving serious and harmful crimes.  (Castaldi involved a big Ponzi scheme impacting many persons, Santiago-Rivera involved a police officer shooting.)  In addition, both rulings show that circuits may be especially inclined to find way-above-guideline sentences reasonable, even more so than way-below-guideline sentences.  (I cannot readily think of one, let alone two, sentences set a full decade below the calculated guideline sentence upheld after an appeal by the government.)  But, as the title of this post suggests, perhaps the broader story is how very few federal sentences are even found unreasonable.

Booker has been the federal sentencing law of the land since 2005, but the true era of modern reasonableness review likely should be defined as starting in December 2007 after the Supreme Court handed down the last of the troika of reasonableness review cases via Gall, Kimbrough and Rita.  Since those opinions were issued, we have probably had over 500,000 federal sentences imposed, and I suspect that less than 250 have been found procedurally unreasonable on appeal and less than 50 have been found substantively unreasonable.  (This federal defender document titled Appellate Decisions After Gall (and updated through early December 2013) provides the most complete accounting of reasonableness rulings that I have seen.)

In other words, based on this very rough assessment of reasonableness review outcomes compared to sentences imposed, it would seem that in only about one of every 2,000 federal sentencings does something go procedurally wrong and in only one of every 10,000 federal sentencings involves some substantive unreasonableness.  (Of course, the vast majority of federal sentences are not appealed, in part due to the wide use of appeal waivers in plea agreements, so the outcomes of appeals is not the ideal measuring stick here.  Still, I think these numbers are telling.)

For a lot of reasons, the prospect of reasonableness review may do a lot more work and have a lot more influence than the so-very-rare reversal of a sentence as unreasonable would suggest.  Still, I have largely given up my prior habit of regularly report notable federal circuit rulings concerning reasonableness appeals in part because affirmances in cases like Castaldi and Santiago-Rivera are, statistically speaking, not really notable.

UPDATE: I just noticed that a Fourth Circuit panel also issued an opinion on the same day as the First Circuit opinion linked above that affirms as reasonable a sentence set about 8 years above the guideline range in US v. Washington, No. 13-4132 (4th Cir. Feb. 28, 2014) (available here).

March 2, 2014 at 07:55 PM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Has anyone formally calculated exactly how very few federal sentences are found unreasonable?:


Three thoughts:

1) If the law of sentencing is sufficiently developed, rates of reversal should be rare. An intelligent judge will know the limits of her authority and will sentence a defendant within the range that is sufficiently justified on the record to stand up on appeal.

2) Even if the law of sentencing is not well-developed, the possibility for reversal will depend, in part, on the breadth of the sentencing court's discretion. If you give trial court's great freedom to go above and below the guidelines, few sentences will be outside the scope of that discretion. If on the other hand, you require a significant justification to depart from the guidelines, more sentences will be struck down.

3) Regardless of the scope of discretion, the potential for reversal will depend on how well the record is developed in trial courts to support departures from the guidelines and how much judges are departing from the guidelines. If most judges are staying within 10% of the guideline range on 99% of the cases and only departing if a strong record is made, there will be few reversals. If judges are ignoring the guidelines entirely and departing by 50% or more in the majority of their cases, there will be more reversals.

In my mind, at some point, what is needed -- maybe in another 3 o4 4 years when the data pool is even larger will be some detailed analysis of how much judges are departing from the guidelines, when they are departing (i.e. what types of facts in the record seem to get judges to depart -- mental health problem, history of good works, number of victims), and when the appellate courts are finding the departure unreasonable.

Posted by: tmm | Mar 3, 2014 9:42:34 AM

The sample pool for appellate review of above-guideline sentences is different from that for below guideline sentences. The federal government very rarely appeals -- where it does appeal, it's likely that the case has not just a below-guideline sentence but also other factors making the sentence imposed seem exceptional. Defendants, by contrast, are likely.to appeal nearly every above-guideline case -- so many of those will be for less drastic deviations.

Posted by: Analysis | Mar 3, 2014 11:26:43 AM

Out here in the sagebrush, the Tenth Circuit has reversed exactly one sentence as substantively unreasonable - unreasonably low of course. They have never, ever, not once reversed a within or above-guideline sentence as substantively unreasonable.

I wish the Tenth Circuit would just declare that there is no such thing as substantive-reasonableness review of within or above-guideline sentences and stop the charade. It would save a lot of time and resources.

Posted by: lawyer | Mar 3, 2014 12:33:01 PM

SCOTUS told us years ago in Booker, Gall and Kimbrough that district courts could get away with almost anything. Why is anyone surprised that the courts of appeals got the message?

Sure, for gross and scandalous outliers, like "sentences" that consisted of blowing a kiss to a traitor (Lynne Stewart) or writing a love letter to a repeat distributor of child pornography (Corey Reingold), district courts can get reversed. But mostly anything-goes sentencing is still mostly anything-goes.

Posted by: Bill Otis | Mar 3, 2014 2:05:33 PM

i'm trying to figure out how the vast vast majority of sentences (am i crazy to believe within-guideline sentences account for over 90% of all sentences? please correct me! = within guideline ranges = anything-goes sentencing. it's just so weird that there are people who really make that argument with a straight face. but there you have it. what do cases of non-gross, non-scandalous outlier sentences look like? what are you complaining about? too many bottom-of-the-guideline sentences? it's really odd.

Posted by: anonymous | Mar 3, 2014 5:03:25 PM

anonymous --

"am i crazy to believe within-guideline sentences account for over 90% of all sentences?"

I don't know if you're crazy but you're certainly wrong. About 52% to 53% of sentences are within range. All the rest are below range, except for maybe 1% to 2% that are upward departures.

The rest of your comment is too incoherent to try to formulate an answer.

Posted by: Bill Otis | Mar 3, 2014 5:41:33 PM

Actually, of the 45.6% of sentences below the guideline range in fiscal year 2012, more than half (60%) were government-sponsored. Only 17.8% of sentences were below the guidelines as a result of judicial discretion, and a good proportion of those were agreed to by the government, either expressly or by not objecting. So it's not "anything goes." It's "73.4% within-guideline range or government-sponsored below-guideline sentences goes."

Posted by: A defender (of humanity) | Mar 5, 2014 12:25:20 PM

A defender (of crooks) --

Thank you for verifying that, as I said, about 52% to 53% of sentences are within range. All the rest are below range, except for maybe 1% to 2% that are upward departures.

Sure, out of the leniency you spend your professional life seeking, and out of wanting to placate the judge, prosecutors often agree, or are bullied into agreeing, to a chunk of the downward departures. I never said otherwise, and the figures I gave are correct.

Posted by: Bill Otis | Mar 11, 2014 5:32:41 AM

In Richmond, Virginia, federal judges often impose above-range sentences in illegal reentry cases. In one case, the defendant received a stat-max sentence of 20 years. See United States v. Rivera-Santana, 668 F.3d 95 (4th Cir. 2012). Granted, the defendant in that case had a significant criminal history. In my experience, however, EDVa judges in Richmond are varying upward in typical 1326 cases, as well. The Fourth Circuit has found all of the sentences the FPD has appealed to be reasonable.

Posted by: FPD lawyer | Mar 13, 2014 11:47:36 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