« Where the jobs (and future sentencing issues) can be found | Main | Eleventh Circuit to review en banc below-guideline sentence in "utterly gruesome" sex offense case »

August 12, 2009

Notable new paper with interesting post-Booker data analysis

This new paper on SSRN by Ryan Scott, which is titled "In Search of the Booker Revolution," presents new data and analysis on how one district is sentencing after Booker.  Here is the abstract:

In 2005, the Supreme Court in United States v. Booker rendered the United States Sentencing Guidelines advisory.  Arriving after eighteen years of complex and mandatory sentencing rules, the decision initially was heralded as revolutionary, both by critics and defenders of the federal Guidelines. But subsequent reports by the Sentencing Commission have shown few signs of a Booker revolution, revealing surprisingly modest changes.  The existing research on post-Booker sentencing is incomplete, however, because it has not examined the response of individual judges to the decision.  That omission is critical, given that the reduction of inter-judge disparity was the central purpose of the Guidelines. Studying sentencing patterns by individual judges is notoriously difficult because the Commission does not disclose the identity of the sentencing judge when releasing case records.  But one district court — the District of Massachusetts — has adopted a unique policy that makes key sentencing documents available to the public, allowing the analysis of judge-specific data.

This Article offers the first empirical evidence of individual judges’ responses to Booker, drawing on a dataset of sentences from 2002 to 2008 by judges in Boston who share a common case pool.  An analysis of those sentences suggests a modest but clear increase in inter-judge disparity since Booker.  The strength of the relationship between the identity of the sentencing judge and sentence length has increased, by some measures 60-80% above pre-Booker levels.  The identity of the judge also has become a stronger predictor of sentencing relative to the Guidelines: some judges now sentence outside the guideline range more frequently, and to a greater extent, than their colleagues.  Although it is difficult to know whether similar trends have played out in other districts, the Boston data suggest that judges’ disagreements about the Guidelines have a greater effect on sentencing outcomes since Booker

This interesting and important data provides a vakuable new empirical perspective on Booker's impact, though the District of Massachusetts is not a perfectly representative sentencing district for lots of reasons.  Still, for everyone eager to get a handle on Booker's impact in at least one district, this paper provides the latest, greatest view from the sentencing courts.

August 12, 2009 at 11:20 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Notable new paper with interesting post-Booker data analysis:


Give the allegedly low levels of inter-judge disparity pre-Booker (i.e. a low base number), the 60-80% increase in inter-judge disparity levels may still be pretty modest.

I look forward to seeing the deviaton from guideline rates in the artice and in particular, the ratio of upward to downward deviation.

Posted by: ohwilleke | Aug 12, 2009 2:31:37 PM

Nationwide, within range sentences have decreased from 70.5% in first quarter 2005 (pre-Booker) to 57.7% in second quarter 2009. An increase in below-range sentences comprises the great bulk of the change with below-range sentences going from approximately 5% to more than 16% during the same period.

Inter judge disparity has always been a problem and future research will likely show the chasm has deepened.

Posted by: mjs | Aug 12, 2009 2:50:52 PM

D. Mass. contains Judge Young and Judge Gertner. Anyone who's followed the blog for a while might understand why D. Mass. might not be a perfectly representative sentencing district. Agree or disagree with their approaches, they've both given a lot of thought to sentencing and spilled a lot of ink over it, and are probably outliers on both scores.

See, e.g.,



I'll be interested to read Prof. Scott's article.

Posted by: anonymous | Aug 12, 2009 2:56:33 PM

Nice study about the obvious.

Posted by: SheliaB | Aug 12, 2009 6:57:23 PM

The more important question is whether crime victimization has increased in the jurisdiction. It would do so by 1) reducing sentences, from mostly downward deviance from guidelines (less incapacitation); 2) deterring prosecutors into offering better plea deals, that also increase the number of criminals in circulation (tough to measure cognitive deterrence of prosecutors - average prison sentences may not reflect the higher fraction that gets no prison time); 3) encouraging criminals with lighter sentences overall (less risk for crime, from its already nearly infinitesmally small risk).

These victimization measures are not available in the paper. If found elsewhere, they should be repeated in another 5 years, since it takes 10 years for a law change to really take effect.

Posted by: Supremacy Claus | Aug 12, 2009 8:45:15 PM

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