Judge Alex Kozinski Website
Alex Kozinski, Judge, United States Court of Appeals for the Ninth Circuit
Once or twice a year I sit as a trial judge, usually in a criminal case. I've been doing it for many years and never cease to learn something new. I started before the Sentencing Guidelines, so I have sentenced defendants both with and without the Commission's guidance. I've had my doubts about the Guidelines; I even wrote an opinion saying they were unconstitutional. n1 Doubts sometimes resurface when I sit as an appellate judge, n2 but I have no doubts when I sit as a trial judge: I like the Guidelines and hope they're here to stay.
I found sentencing traumatic in the pre-Guidelines days. The sentencing range often spanned many years, sometimes all the way from probation to life in prison. Some judges may have the wisdom of Solomon in figuring out where in that range to select just the right sentence, but I certainly don't. Would too heavy a sentence destroy a young life and snuff out any chance of rehabilitation? Would too light a sentence embolden the defendant, endangering the lives of innocents? What deterrent effect will the sentence I impose have on others? n3
Deciding whether someone spent the next twenty years in prison or got straight probation was a burden almost too heavy to bear. Somehow I felt it was wrong for one human being to have that much power over another. Imposing sentence was, for me, almost an act of sacrilege.
Nor did it help that there was no appeal from the sentence. n4 On the contrary, the very finality of the decision made it all the more difficult. If the sentence had been subject to appeal, someone else could have told me if I was way off the mark. But with no sentencing appeals, the burden lay entirely on my shoulders, and I didn't like it a bit.
Enter the Sentencing Guidelines and all this changed for the better. Gone are the wide open spaces for sentencing discretion. Sentencing ranges are narrow and presumably take into account all those factors I don't feel competent to weigh: punishment, deterrence, rehabilitation, harm to society, contrition--they're all engineered into the machine; all I have to do is wind the key. The Probation Officers in the Central District of California, where I mostly sit, do an excellent job and there are seldom disputes over the accuracy of the PSR. When there are, I do what I am competent to do--make factual findings.
Once I have figured out the range, I always sentence at the very bottom; I never depart up or down, unless it's a guided departure like substantial assistance or acceptance of responsibility. This is true whether a defendant has pleaded guilty or proceeded to trial; generally, I have found that the bottom end of a given Guideline range sufficiently captures a defendant's criminal culpability, and I very seldom run across a case so unusual as to warrant departure. If the sentence seems too harsh or too light, I no longer feel responsible.
Alas, it was not to last. I have noticed a significant increase in requests for departure--particularly downward departure. It seems anything and everything that might make the crime or the defendant different from the norm can trigger a request. As requests increase, so have departures. As early as 1991, research indicated that "in a significant minority of cases, departure is driven by the sentencing judge's desire to reach a result different from that specified in the Guidelines, rather than by the presence of meaningfully atypical facts." n5 It is my perception that this practice has become ever more common as appellate review of departures has been significantly narrowed in the wake of Koon v. United States, 518 U.S. 81 (1996). n6
While Koon has limiting language, particularly the heartland discussion, other parts of the Court's opinion can be read as giving district judges broad and unreviewable discretion to depart. n7 Our circuit has even held--in an opinion about which I have since developed serious doubts--that Koon broadens not only the authority to depart, but the extent of departure as well. n8
Koon seems to have jammed a crowbar into the Sentencing Guidelines machinery. Departures are coming to be the norm rather than the exception. As of 1998, downward departures had increased for eight years in a row, n9 and in some districts have become quite common. In the District of Arizona, for example, 61% of sentences included a downward departure that was based on grounds other than substantial assistance, n10 I suppose that every case has its own unique claim to atypicality but, in the aggregate, I find it hard to believe that more than half the cases in a district are atypical. It looks to me like we are inching toward a system very similar to the one in existence prior to the Guidelines, with broad sentencing discretion vested in the district court, and relatively little appellate review.
The incoming Commissioners might want to consider whether the frequency with which departures are now being granted by district courts is consistent with the basic premise of consistency and uniformity which is supposed to be the backbone of the Sentencing Guidelines. Or, to put it another way, if we're going to have wide sentencing disparities anyway, what's the point of keeping the Sentencing Guidelines and a Sentencing Commission?
In my view, Koon went too far. In attempting to give district judges some wiggle room to depart in extraordinary cases, Koon gave away the store. The problem is that there may be no way to announce a legal standard that separates the truly extraordinary case from other cases that are merely unusual. It is a question of judgment, and the exercise of judgment always carries with it the risk of abuse.
While no one tinkers lightly with a Supreme Court opinion, Koon was an interpretation of the Guidelines and therefore subject to reconsideration by the Commission. Cato the Elder used to finish every speech with his famous exhortation about the destruction of Carthage. It worked for him, so I will end my own disquistion with a similar call for action: Koon must be overruled!
n1 See Gubiensio-Ortiz v Kanahele, 857 F.2d 1245 (9th Cir. 1988), wrongly overruled by Mistretta v. United States, 488 U.S. 361 (1989).
n2 I wonder, for example, why it is OK under certain circumstances to sentence a defendant based on uncharged conduct, or based on charges dropped as part of a plea bargain. See, e.g., United States v. Fine, 975 F.2d 596, 602-03 (9th Cir. 1992) (conduct dismissed or uncharged pursuant to a plea agreement can be relevant for sentencing purposes); United States v. Loveday, 922 F.2d 1411, 1417 (9th Cir. 1991) (upward departure based on conduct in dismissed counts permissible because conduct was relevant to offense of conviction).
n3 See Alex Kozinski, Teetering on the High Wire, 68 U COLO. L. REV. 1217, 1218-20 (1997).
n4 See, e.g., United States v. Tucker, 404 U.S. 443, 447 (1972) ("[A] sentence imposed by a federal district judge, if within statutory limits, is generally not subject to review.").
n5 Michael S Gelacak et al., Departures Under the Federal Sentencing Guidelines: An Empirical and Jurisprudential Analysis, 81 MINN. L. REV. 299, 364 (1996) (emphasis omitted) (summarizing review of over 1,400 district court departure decisions).
n6 See, e.g., United States v. Lipman, 133 F.3d 726, 729-30 (9th Cir. 1998) (holding that district court could consider departure based on "cultural assimilation," but decision not to depart was unreviewable on appeal).
n7 See, e.g., Koon, 518 U.S. at 97 (emphasizing limited nature of appellate review and traditional deference given to sentencing court); id. at 98 (noting district court's "institutional advantage" in determining cases that justify departure because of unusual factors); see also United States v. Mendoza, 121 F.3d 510, 515 (9th Cir. 1997) (noting that determination of "what falls within the 'heartland' of a Guideline is ... within the discretion and special expertise of the district court") (quoting Koon, 518 U.S. at 96).
n8 See United States v Sablan, 114 F.3d 913, 918-19 (9th Cir. 1997) (en banc).
n9 See US. SENTENCING COMMISSION, 1998 ANNUAL REPORT 39 (1999).
n10 See id.