After William Kirkpatrick was arrested for possessing a gun — something his felony-convictions made unlawful, see 18 U.S.C. § 922(g)(1) — he confessed to committing four drug-related murders. He also told a cellmate that he had arranged a contract hit on the federal agent who was conducting the investigation. After investing more than 200 hours in investigations, agents concluded that Kirkpatrick had neither killed anyone nor arranged for the agent’s death. Kirkpatrick pleaded guilty to the felon-in-possession charge. The Sentencing Guidelines recommended a range of 37 to 46 months. The judge gave him more — sensibly so. Lying to a federal agent is a crime, see 18 U.S.C. § 1001, and so is threatening the life of a federal agent, see 18 U.S.C. § 115(a)(1)(B). Enhancements on account of an offender’s additional crimes are normal and proper. See
United States v. Watts,
Kirkpatrick contests not the fact of the enhancement but its magnitude. The district court imposed a sentence of 108 months’ imprisonment, more than double the top of the Guidelines’ range. Kirkpatrick calls this unreasonable: an extra five years in prison is an exceptionally harsh sentence for wasting 200 hours of federal employees’ time. Whenever a court gives a sentence substantially different from the Guidelines’ range, it risks creating unwarranted sentencing disparities, in violation of 18 U.S.C. § 3553(a)(6), for most other judges will give sentences closer to the norm. That’s a major reason why substantial variances from the Sentencing Commission’s recommendations require careful thought. See
Gall v. United States,
When a variance is carefully explained, appellate review is deferential. But when a sentence appears to be chosen arbitrarily, it is problematic. That is the case here. The judge said that Kirkpatrick’s lies, which put the case agent in fear and wasted the time of agents who had better things to do, justified a higher sentence. That cannot be gainsaid. But five years’ extra time in prison? The judge proceeded as if any sentence within the statutory maximum (10 years) needs no explanation beyond the conclusion that something more than the top of the Guidelines’ range is in order. Yet every sentence must be justified under the criteria in § 3553(a), and the district judge does not appear to have recognized that leaping close to the statutory maximum creates a risk of unwarranted disparity with how similar offenders fare elsewhere — not only because this may overpunish braggadocio, but also because it leaves little room for the marginal deterrence of persons whose additional deeds are more serious (for example, actually putting out a contract on an agent’s life).
Before
United States v. Booker,
Kirkpatrick’s range of 37 to 46 months comes from his offense level of 20 and criminal history category of II. Suppose his lying were treated as incompatible with accepting responsibility, costing Kirkpatrick the two-level reduction he had received. That would produce a total of 22 and a recommended range of 46 to 57 months. (Kirkpatrick’s offense level of 20 already included an enhancement for obstructing justice.) If Kirkpatrick also were charged with, and convicted of, lying to federal agents and making threats designed to intimidate agents, the offense level could reach 24 (depending on how the grouping rule, U.S.S.G. § 3D1.2, treated the § 1001 and § 115 convictions, which have lower base levels than the felon-in-possession crime). At offense level 24 and criminal history category II, Kirkpatrick’s recommended range would have been 57 to 71 months. There are some other ways of classifying his additional acts that might produce higher ranges, but level 24 seems most likely. To get as high as level 28 (which corresponds to a range of 87 to 108 months), the district judge would have to find that Kirkpatrick actually set out to have the case agent murdered. Given the judge’s belief that Kirkpatrick was all bark and no bite, however, the Sentencing Commission’s recommendation is for a sentence substantially below 108 months.
Booker
and its successors mean that this recommendation is not conclusive. But before exercising discretion the judge should know what that recommendation is, and thus how Kirkpatrick’s sentence will compare with the punishment of similar persons elsewhere. The Supreme Court has never questioned the principal goal of the Sentencing Reform Act of 1984: to curtail the variable sentencing caused by different judges’ perceptions of the same criminal conduct. The allowable band of variance is greater after
Booker
than before, but intellectual discipline remains vital. “[A] motion to [a court’s] discretion is a motion, not to its inclination, but to its judgment; and its judgment is to be guided by sound legal principles.”
United States v. Burr,
