Trying to parse the USSC's latest data

The latest post-Booker sentencing data released by the US Sentencing Commission (discussed here and here) are fascinating.  But I find it extraordinarily challenging to draw any firm conclusions about exactly what the data all means.  Here are a few quick reactions after having a little time to look over the numbers.

No drifting?:  Especially because average and median sentence lengths seem stable (or even rising a bit) post-Booker, it is hard to find a lot of support for the contention made by AG Alberto Gonzales, in his speech last month advocating a Booker fix (basics here, commentary here and here and here), that we are seeing a "drift toward lesser sentences."  And yet, because there are more below guideline sentences now than pre-Booker, perhaps the sentence length data reflect a change in the mix of cases.  (Indeed, I have been speculating since Blakely that fewer cases might get prosecuted in the federal system due to all the legal commotion, and that the least serious cases would be those most likely to be left to the states.)

Increased disparity?: In his speech last month, AG Gonzales also asserted that "the evidence the Department has seen since the Booker decision suggests an increasing disparity in sentences."  The USSC's latest numbers perhaps provide support for this contention based on the circuit-by-circuit data.  The rates of judicial variation from the guidelines are distinctly different in different circuits (e.g., in the Second Circuit, judges are initiating departures or Booker variances in roughly 1 of every 4 cases, while in the Fifth Circuit, judges are doing so in roughly 1 of every 10 cases).  But, of course, the rates of prosecutorial variation from the guidelines are also distinctly different in different circuits (e.g., in the Ninth Circuit, prosecutors initiate departures in roughly 1 of every 3 cases, while in the First Circuit prosecutors do so in roughly 1 of every 8 cases).  And, as pre-Booker data on departures from 2002 and from 2003 spotlight, significant circuit-by-circuit variations were common in the pre-Booker era.

So, what firm conclusions can be drawn?  Perhaps only that the Booker experiment has not (yet?) produced a radical change in federal sentencing outcomes, and thus there does not seem to be a dire need for the sort of Booker fix urged by AG Gonzales.  Because, as stressed in this post and in my USSC testimony here and here, another round of legal confusion and uncertainty would likely follow any major structural changes to the federal sentencing guidelines in the wake of Booker, I think Congress would be wise to leave well enough alone at least for the time being.