So You Want to Structure Sentencing? Part V

Special Cases: Part V takes on the special case of Re-structuring a sentencing process that is widely viewed (by judges especially) as in need of reform, and attempts to knit together some points from previous blogs.

Previous posts in this series have largely started from the point of view of policy makers considering structured sentencing for an indeterminate sentencing system, by definition loosely structured with wide discretion for judges and corrections officials. But what about structured sentencing systems that are considering wholesale re-structuring? What special problems occur? Let’s take the examples of the federal U.S. guidelines and California (I am not suggesting that there is agreement among all parties favoring re-structuring, only that it is widely discussed).

California. We have discussed the challenge facing California (See California challenge series). California’s structure is a form of legislatively-enacted guidelines. Roger Warren, in his comments before the Ca. Little Hoover Commission (here) discusses the most pressing need in the area of correctional reform IS sentencing reform, including a reform of the complex, “Byzantine” criminal code and other problems that hamper judges. I might add that a cursory look at the California statutes suggests to me that judicial discretion to tailor sentences is sharply limited, leaving little room for consideration of, for example, offender prior record as a predictor of future public safety risk (See also Tom McGee’s comments on Part IV of this series and elsewhere – thanks Tom, for presenting an interesting view and proposed general remedy). Of course, the exception that proves the rule is the draconian 3-strikes law which brings the house down on the heads of selected defendants so convicted.

Federal guidelines. The latest issue of the Federal Sentencing Reporter (intro by Frank Bowman here) takes up the Constitution Project’s Sentencing Initiative (CPSI) and its call for reform of the federal sentencing guidelines. In particular, the CPSI calls for simplification and flexibility, finding the federal guidelines too complex and inflexible in comparison to other guideline systems. Bowman outlines the working group’s criticism of the existing system and its proposed remedy. Clearly a thoughtful group whatever your position on the issue.

Yet reforms (or re-reforms, rather) in both California and the federal system face major obstacles. Recent posts have briefly described the obstacles to reform in California (The California Challenge). Regular readers of Doug Berman’s weblog are well versed in the obstacles to the kind of substantive change in the U.S. federal guidelines that CPSI proposes, and the forces that want to re-impose features of the federal guidelines that CPSI rejects.

The Wisdom of Crowds.
I believe that one of the reasons for the intractability of these problems relates to a frequent rant of ours, decent baseline data on past practice (Part I and IV of this series, etc.). Decent baseline data gives system participants much confidence that the reform is grounded on past practice, offering selective changes while minimizing unintended consequences and chaos. “Hold On!” you say. California and the USSC have excellent data, at least on sentences. But therein lies the rub!

The recent historical data on California and federal sentencing practice reflects data on sentences that were TIGHTLY CONSTRAINED in actual practice, and do not reflect what judges believed was the right thing to do in at least a subset of problematic cases (or perhaps whole classes of cases), rather what they were required to do. Extending the recent post, Sentencing and the Wisdom of Crowds, the wise crowd in this case is the large body of judges, and what they thought a fair sentence under a discrete set of circumstances was. The existing tightly-constrained baselines for the USSC and California are no baseline at all to tell you what the “crowd” thought, only what the crowd was compelled to do.

So what do you use for a baseline (if you're still on board/ever were on board)? The Wisdom of Crowds post, below, argues (rightly, in my judgment) that we have plenty of useful data nationwide on which to build, but California and the U.S. federal systems are exceptions, if you buy my logic. The baselines that exist there are nearly meaningless as a starting point for substantial reform, skewed by tight constraints.

What are the options? (I could use your help here – easier to diagnose than to treat)
(1) Start from scratch. In essence, blow up the existing system, and go back to a highly discretionary system for at least a couple of years, allowing the judges to speak with unfettered voices. This solution, which best captures the “wisdom of crowds,” looks implausible politically. How can we expect policy makers to admit so publicly their doubts about the system that they and their predecessors devised? And probably many do not harbor such doubts.
(2) Conduct thought experiments and sentencing councils. Collect experts (judges, for example) and use short hypothetical cases (vignettes of crimes) to discover the group mean sentence without reference to the tightly constrained structure. Sophisticated variations on this approach have been used with a random sample of citizens too (US Sentencing Commission’s 1997 Just Punishment study) and recent work by Mark Cohen, et.al. (here). However, as I previously argued in Part I of this series, it is very difficult to simulate actual circumstances and actual decisions, and is probably no substitute for actual historical sentences.
(3) Proceed incrementally. Probably the most likely approach, also flawed given the difficulty of making landmark substantive change incrementally. Pick your battles. Look for evidence to support structural and procedural change (e.g., problems caused by undue complexity), and evidence to make piecemeal change to particularly problematic crime categories etc. But if the structure is the problem, how far will incrementalism really take you?
(4) I’m out of ideas – how about you?

The bottom line: Sentencing structures have substantial inertia, and re-structuring a system that is 20-30 years old and has already constrained earlier forms of substantial discretion may be harder than structuring as age-old system that appears “unreformed.” The old line “If it ain’t broke, don’t fix it” will be echoing through the legislative corridors – and an ironclad case for “yes, broke” is almost surely necessary.