On Friday, the Seventh Circuit ruled in United States v. Scott (No. 10-1597) that a judge sentencing one conspirator may not take into account the government’s failure to prosecute a coconspirator. Here’s what happened. In 2009, a federal grand jury charged Scott and a coconspirator with mail fraud. Scott pled guilty and was sentenced to 63 months in prison, while, for unknown reasons, the government dismissed all charges against the coconspirator. On appeal, Scott argued that the sentencing judge erred by not taking into account the unexplained dismissal of charges against the coconspirator. In essence, his theory was this: federal sentencing law is intended to minimize unwarranted disparities as between similarly situated offenders, and the disparity between one conspirator getting a long prison term and the other one not even getting a conviction is a very profound one indeed.
The Seventh Circuit didn’t bite, holding categorically that “we reject the notion that a court may consider the coconspirator’s lack of a conviction under § 3553(a) because holding otherwise would deprive prosecutors of the opportunity to exercise any meaningful prosecutorial discretion in coconspirator cases” (8).
I agree, however, with Judge Williams’ concurring opinion that such a broad holding was unnecessary and regrettable.
It is one thing to say that a sentencing judge is not required to consider the coconspirator — which would have been enough to dispose of Scott’s appeal — and quite another to say that a sentencing judge is not permitted to do so. In this post-Booker world, federal sentencing is controlled by § 3553(a), which on its face grants broad discretion to sentencing judges to take into account a wide range of broadly worded considerations (e.g., “to promote respect for the law, and to provide just punishment for the offense”). Given the scope of post-Booker discretion, I would think it generally inappropriate for an appellate court to go out of its way to rule out a potential sentencing factor on a categorical basis.
To be sure, the particular factor that Scott was pushing strikes me as one that will rarely merit much weight. Standing alone, the argument that “I got X years, and he got nothing” does not speak in a clear way to whether the sentence was just — on the face of things, it seems about as likely that the coconspirator received unwarranted lenience as that the defendant received unwarranted harshness. Why give the defendant a break based on the coconspirator’s windfall?
Still, I could imagine this sort of argument adding strength to other arguments. If there were good reasons to think that a guidelines sentence was unwarranted in a given case, the fact that prosecutors did not seek any punishment at all in the case of a coconspirator might help to confirm a preliminary judgment that lenience was appropriate. Moreover, taking the disparity into account would help to address the defendant’s perception that the system operates in an arbitrary fashion — bearing in mind (as I demonstrate in this article) that the original intent of federal sentencing reform included a desire to diminish the feelings of resentment that defendants experienced as a result of arbitrary treatment.
What really seems to drive the broad holding in Scott (as well as similar holdings in cases that present related issues, like the disparate treatment of defendants in federal and state court) is a separation of powers concern. Recall the language quoted above: “holding otherwise would deprive prosecutors of the opportunity to exercise any meaningful prosecutorial discretion in coconspirator cases.”
I attempted to rebut this sort of reasoning several years ago in this article, and I must say I continue to be mystified by the suggestion that anything a judge does at sentencing intrudes impermissibly into the sphere of prosecutorial discretion. The sentencing judge does not — cannot, really — undo a prosecutor’s decision not to prosecute. Nor can the sentencing judge do anything to alter the conviction that the prosecutor has worked so hard to obtain. All the judge can do is to decide what degree and type of punishment should be associated with that conviction. This is a core, traditional function of judges, not prosecutors. The legislature sets the parameters within which this sentencing discretion is exercised, and could rule out consideration of what prosecutors have done in other cases, but the legislature has not done so.
To be sure, looking to the treatment of other defendants would lead judges into a position of second-guessing and sometimes expressly criticizing prosecutors’ decisions, but the criticism by an official of one branch of the work of an official of another branch does not present a separation of powers problem — judges take plenty of heat themselves on a regular basis from the executive and legislative branches.
The question is, or should be, whether the work of one branch is interfering in a significant way with the core functions of another branch. The fact that a prosecutor may end up being disappointed in the sentence imposed in one case does not undermine the prosecutor’s ability to decline, charge, plea-bargain, or try other cases, except perhaps in the very marginal way that lenience in one case may diminish plea-bargaining leverage in other cases. But this loss of leverage results whenever a sentencing judge is lenient for any reason; it is a not a problem specific to taking into account outcomes in other cases.
At bottom, for a sentencing judge to do the sort of thing that Scott requested seems like a far less serious intrusion into the sphere of prosecutorial discretion than is the judge’s well-established, unquestioned power to reject a plea deal favored by the government.
As for the curious suggestion, made in Scott and many other cases, that judges lack the competence to evaluate prosecutorial charging decisions, see my post here.