« Weekend reading (as I head back to off-line land) | Main | Lots of lethal injection talk »

May 7, 2006

Policy judgments at federal sentencing: aren't they inevitable and mandated by Congress?

In discussion of post-Booker federal sentencing at the Yale Law School class I recently had the pleasure of attending, I was surprised to often hear the refrain that judges "should not make policy judgments" at sentencing.  I have long thought it was post-modern gospel that all judicial decision are, at some level, policy judgments.  But beyond this legal realist premise, it dawns on me that all judicial sentencing decisions plainly are, at some level, policy judgments informed by views on just punishment, crime control, procedural fairness, and other express and implicit considerations.

After Booker, the impact of policy judgments at sentencing is evident in many ways.  Decisions by some district judges to give the guidelines heavy weight obviously reflect a kind of policy judgment, as do decisions to apply a particular burden of proof or to increase a sentence based on acquitted conduct. Likewise, decisions by circuit judges to handle Booker plain error in particular (and diverse) ways reflect a policy judgment, as did the Supreme Court's decision not to resolve the circuit split over Booker plain error standards.

As we reflect on these realities, it is valuable to appreciate that executive branch officials make policy judgments at sentencing all the time.  All prosecutorial charging and plea bargaining decisions involve, at some level, policy judgments informed by views on just punishment, crime control, procedural fairness, and a range of other express and implicit considerations.  Same goes, of course, for decisions about who gets sentence reductions for assisting authorities or which districts will have fast-track sentence reduction programs.  And, interestingly, we never hear complaints or concerns about unelected federal prosecutors making policy judgments at sentencing (even though prosecutorial policy judgments are not made in open court nor subject to any kind of review).

Furthermore, it seems that Congress, through the text of 3553(a), has told federal judges to make individualized policy judgments at sentencing.  The formal mandate of 3553(a) requires a federal judge to impose a sentence in each case that is "sufficient, but not greater than necessary" to serve the traditional purposes of punishment set forth in 3553(a)(2).  Doesn't this text suggest that Congress wants federal judges, in individual cases, to make policy judgments about how the traditional purposes of punishment ought to be applied to individual defendants?

In sum, I questions the assertion that federal judges "should not make policy judgments" at sentencing because: (1) such judgments seem inevitable and are clearly made by prosecutors in every criminal case, and (2) a fair reading of the text of 3553(a) suggests that Congress has ordered federal judges to make individualized policy judgments in each and every sentencing.

May 7, 2006 at 04:51 PM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Policy judgments at federal sentencing: aren't they inevitable and mandated by Congress?:


Thanks for some basic clarity on this topic. I'm no expert in criminal law, or any kind of law for that matter, but from the outside looking in it seems rather transparent that judges routinely make 'policy judgments' at sentencing, irrespective of one's thoughts about 'legal realism' or 'post-modernism.' While law and politics (or 'public policy') are not co-extensive, it is clear that sentencing presupposes and assumes any number of policy orientations, and this accords as well with Lon Fuller's claim that law in general incarnates both procedural and substantive values. I'm curious what others might think are the the reasons (implausible or otherwise) that some would dare to argue that 'judges should not make policy judgments,' as this admonition seems otherwise steeped in appalling ignorance or naivete. Again, thanks for some fundamentally sound reflections on this subject.

Posted by: Patrick S. O'Donnell | May 8, 2006 9:55:19 AM

I've argued that, within the guideline range, judges can consider virtually anything, policy or not. Williams v. New York, 337 U.S. 241, 246-51 (1949); Witte v. United States, 515 U.S. 389, 398 (1996) (even under guidelines, factors that could be considered within the range were “largely unlimited”). The only restriction on factors that could be considered on setting a sentence within the legal range has been a prohibition on considering impermissible factors such as race and nationality. Courts could always consider "policies" such as the need, or lack of need for deterrence, in setting a sentence, and can now consider anything that is relevant to the factors in § 3553(a).

Posted by: David Lewis | May 8, 2006 10:10:59 AM

Ditto to Patrick O's comments.

I'm curious: for those individuals who say that judges "shouldn't make policy choices at sentencing," what do you thinks motivates this thinking? I don't even understand the claim (basic 1L lesson -- at common law and in equity, courts make policy judgments), so I'm curious what motivates it. Also, what is the reputation of people who say such a thing? You're there, so maybe you can tell us.


Posted by: Mark | May 8, 2006 10:14:48 AM

As someone with a policy background who has been in sentencing for over a decade now, I've been concerned for sometime about the lack of real "policy" perspective in the discussions of sentencing and its impact. This is not a knock on the fine academic work done by legal scholars or criminologists, but the policy academy and practitioners have a lot to say about sentencing that really hasn't been reflected in usual discussions, as much the fault of the latter as the former. It is now a truism in policy literature that the implementors of laws are making policy, plain and simple. The "street-level bureaucracy" literature, launched a couple of decades ago by Michael Lipsky, shows how policy isn't policy until it's in the hands and operation of the people who carry it out, and the concept has been applied well in criminal justice at the law enforcement level by Robert Ker Muir and others. Judges resist the notion because acknowledgement would endanger their status and turf as "beyond politics" and put them on the same plain as the legislators and executives who are trying to get control of the results of what the judges and prosecutors do. Their concern is not unwarranted, but it ignores the fundamental nature of the consequences of their acts. They usually prefer to let the "politicians" deal with those, adopting a "that's not my problem, I just judge cases" attitude when, of course, the results are resources being shifted not only from other policy areas, but also from criminal justice and the courts themselves. The failure to see what they do as policy, in other words, may insulate them from the long-term results of their behavior, but it also threatens to end up having a greater impact on both public safety and judicial discretion than anything they fear from the other policymakers. The remedy for that may be a more aggressive stance from those within the law community for recognition of the policymaking going on, as you have highlighted, and the inclusion of more policy specialists in these forums, who can emphasize both how sentencing is policymaking at the "street level" and how policy models can help practitioners and scholars to better conceive of their common concerns and enterprise. Thanks for bringing this to the wider attention of your audience. It's nice for those of us who see with a policy perspective to have some affirmation once in a while.


Posted by: Mike | May 8, 2006 10:45:57 AM


Those who say judges shouldn't impose sentences based on policy are clinging desperately to the last shards of a mandatory guidelines system. In that mandatory system, a judge was not allowed to consider any factor (often a "policy" factor) that the Sentencing Commission had already considered. Thus, the myth arose that judges were somehow incapable of considering such things -- notwithstanding Congress's direction to the contrary in § 3553(a).

Most particularly, judges were forbidden to consider the rankest hypocricies in the guidelines system -- that sentences varied drastically simply based on the district of prosecution (a factor irrelevant to the factors in § 3553(a)) or that they varied based on factors that sober study has found to be largely baseless, such as the significantly higher sentences imposed for crack cocaine than for powder cocaine offenses.

Prosecutors and some judges simply don't want judges to be able to consider these factors, even though it would rationalize sentencing to a considerable degree if they could. As Doug has pointed out, however, there is now no legal basis for such a restriction.

Posted by: David Lewis | May 8, 2006 10:51:52 AM


Can you please let us know where one can find the argument you refer to above? Is it at SSRN? Where do you teach? (I'm new to this blog)
Thanks so much.

Best wishes,

Posted by: Patrick S. O'Donnell | May 8, 2006 11:38:06 AM

At the risk of appearing "appallingly ignorant or naive," it is not at all clear to me that judges must or should make policy choices at sentencing. This argument seems to confuse judgments that MAKE policy with those that REFLECT or IMPLEMENT policy choices made by others. Of course all sentences reflect policy choices. The statutory mandate Doug cites to impose a sentence "sufficient, but not greater than necessary" to achieve the purposes of sentencing is a polity choice made by Congress that judges are bound to implement.

The tricky point, of course, arises because the list of purposes in the federal statute is exhaustive, the purposes are in tension in some cases, and are stated so generally as to permit a wide variety of different policy choices. Is it up to judges to pick and choose which purposes to pursue on a judge-by-judge and case-by-case basis? If this were so, a "discordant symphony" of disparate sentences in similar cases would result, because judges do not agree about the priorities among purposes. In the pre-Booker era, appellate review did not constrain these choices sufficiently to prevent considerable disparity among judges and regions, and it is doubtful it would do so now.

Given this, the statutory mandate at 3553(a)(6) to "avoid unwarranted sentence disparities" requires judges to strive to implement a COMMON sentencing policy or philosophy. To do this they need to identify the policies of the system as a whole, rather than decide what they personally think should be the policy of the system. It causes a problem that neither Congress nor the Commission has explicitly, consistently, and specifically articulated the policy priorities of the guidelines system. But in many areas of law judges construct or interpret the policies underlying a body of law; they need to do that with federal sentencing law. They key is to try in good faith to implement the system's policies rather the judge's personal views.

This isn't as hard as it sounds, because there is considerable consensus about sentencing policy. Every guideline system in the country implements a version of "limiting retributivism"---which is just a fancy way of saying that the severity of the sentence should reflect the seriousness of the crime, with some consideration also given to the dangerousness of the offender.

A potential problem is that judges will think that implementing the SYSTEM'S policies means rigidly following the guideline rules. This is not necessarily so. The rules are not perfect or complete. They sometimes fail to achieve THEIR OWN PURPOSES. For example, the drug quantity attributable to an offender often reflects the seriousness of the offender's involvement in a drug trafficking crime. But sometimes it doesn't. When the rules don't work right, they should not be applied. Departing in a situation like that isn't substituting one's own policy choices for those of the system. It is implementing the system's policies through individualization to the circumstances of the particular case. That's what discretion is really for.

Posted by: Aion | May 8, 2006 11:51:49 AM

Can you play out your theory with some specifics, Aion? How do you think judges, after Booker, implement policy when dealing with fast-track or crack sentencing issues? Isn't using a 20:1 ratio, and/or considering fast-track disparity, a way of implementing policy in light of the instructions of 3553(a) (and, conversely, wouldn't a circuit decision to categorically reject these concerns perhaps be itself a kind of inappropriate policy-making)?

In short, your distinction might sound nice in theory, but I think it becomes just a matter of sematics if/when we try to operationalize your claims (especially when one also considers that judges have to try to satisfy 3553(a)(2), but only need to consider the concerns of 3553(a)(6)).

Posted by: Doug B. | May 8, 2006 12:06:40 PM

'This argument seems to confuse judgments that MAKE policy with those that REFLECT or IMPLEMENT policy choices made by others.'---I understand the phrase 'making policy judgments' to refer to the fact that in sentencing determination judges are (or should be) reflexively aware of the fact that they are reflecting or implementing policy choices that originate with others. I did not understand it to mean that they were literally making (de novo, or idiosyncratically, or based on personal preferences, etc.) public policy decisions. A particular policy judgment might entail the acknowledgement that there is not, in fact, absolute consensus as to what 'policy as a whole' is, and clarification from the bench as to precisely how the sentence is believed to reflect prior public policy formulations captured in statutory rules and guidelines. I in no way understand the policy judgment being spoken of here as a mandate for the judge to simply implement her personal views of what makes for proper or good public policy. Thanks for further clarifying things here.

Posted by: Patrick S. O'Donnell | May 8, 2006 12:23:07 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