Thoughts on post-Booker data collection

In this post reflecting on last week's House hearings (highlights here), I stressed the importance of focusing upon data rather than anecdote when assessing federal sentencing post-Booker.  But, even as we properly focus on sentencing data post-Booker, we should be mindful of the great Disraeli/Twain quote: "There are three kinds of lies: lies, damned lies and statistics."

The challenge for the US Sentencing Commission and others will be not only to collect and analyze post-Booker sentencing data, but also to present this data accurately and effectively.  I thought the data report on the 733 cases sentenced on or after January 12 as of February 4 in Judge Hinojosa's testimony to the House Subcomittee last week was very well presented, but its accounting of the cases still raised a lot of "coding" questions in my mind. 

Moreover, I fear a lot of coming data confusion in part because we do not even have a settled nomenclature for new-fangled post-Booker sentences that are neither within the guidelines nor represent "old-world" departures: the Second Circuit in Crosby suggested the term "non-Guidelines sentence"; Judge Cassell has coined the term "variance"; the defense bar is partial to "statutory sentence".

On this subject, I see White Collar Crime Prof Blog has this interesting post on related data issues which reports on Judge Nancy Gertner's proposal for developing a "sentencing information system" which seeks to move away from "act of departure" reporting to a "why departing" reporting system.  My casebook co-author Marc Miller should be credited for promoting sentencing information systems in this recent article, which notes their use in Scotland and New South Wales.  Whether and how data could be collected and assembled to create an effective sentencing information system for federal sentencing is one of the many interesting post-Booker issues to watch.