dissenting.
The last time we encountered Kennedy, he raised four challenges to his sentence. United States v. Kennedy, 578 Fed.Appx. 582, 593-96 (6th Cir. 2014). We found three meritless. See id. But we remanded to the district court “for consideration of Kennedy’s argument for a downward variance.” Id. at 596. He had only one downward-variance argument: that his sentence was “unduly harsh” because the Guidelines counted an earlier conviction twice when setting his sentencing range. See id.
On remand, the district court did what we asked: it considered this “double-counting” argument. Ultimately, it rejected the argument and resentenced Kennedy to the same 108-month term. The court stated good reasons for its decision—including that Kennedy had called in a bomb threat to a high school and had a well-document*420ed affinity for illegally possessing and igniting explosives.
Now, however, the majority faults the district court for not addressing Kennedy’s new argument that his post-sentencing rehabilitation efforts should lower his sentence. See Pepper v. United States, 562 U.S. 476, 131 S.Ct. 1229, 179 L.Ed.2d 196 (2011). The majority holds that the district court had to address this argument because we originally issued a “general remand”—a remand which would let sentencing “begin anew.” See United States v. Garcia-Robles, 640 F.3d 159, 166 (6th Cir. 2011). You can tell we issued a general remand, the majority contends, because our previous order lacked an “explicit limitation” on the district court’s review. See Maj. Op. at 414.
I disagree. We remanded “for consideration of Kennedy’s argument for a downward variance." Kennedy, 578 Fed.Appx. at 596 (emphasis added). As used in our order, the word “for” functioned to indicate the remand’s purpose. Why would we state a purpose for remand rather than simply “remanding” or “remanding for re-sentencing”? Simple: to limit the remand by “explicitly outlin[ing] the issue[ ] to be addressed by the district eourt[.]” United States v. Campbell, 168 F.3d 263, 265 (6th Cir. 1999). We had already taken three sentencing issues off the table. We left the lower court to address one—not to start from scratch.
The majority holds otherwise because it contends that our remand order lacked language outlining the “procedure the district court was to follow” and failed to state a “chain of intended events” with sufficient “particularity.” Maj. Op. at 414-15. But given that we found the district court did everything right the first time, save considering the double-counting argument, what else needed to be said? Our order stated the 'only “event” or “procedure” necessary: the district court would consider Kennedy’s double-counting argument.
■ Given the remand order’s scope, we have no basis to reverse here. The court did not have to consider Kennedy’s rehabilitation argument. See Pepper, 562 U.S. at 505 n.17, 131 S.Ct. 1229 (stating that “limited remand orders .;. may render evidence of post-sentencing rehabilitation irrelevant in light of the narrow purposes of- the remand proceeding”). It lacked authority to do so. See United States v. Williams, 522 Fed.Appx. 278, 279 (6th Cir. 2013) (“[Pepper] does not ... empower a district court to exceed the scope of a limited remand.”).
Because I disagree with majority, I feel obliged to address Kennedy’s other challenge: that his sentence is substantively unreasonable. It lacks merit. He simply repeats an argument he made at his first sentencing: that he should receive a downward variance because he made a weak explosive that hurt no one. Kennedy received a sentence within the Guidelines range, however, which makes it presumptively reasonable. See United States v. Mosley, 635 F.3d 859, 865 (6th Cir. 2011). He cannot point to any mandatory sentencing factor the district court ignored. He essentially invites us to “balance the factors differently than the district court did.” United States v. Ely, 468 F.3d 399, 404 (6th Cir. 2006). But we only review for error—not to rebalance the factors in the way we would prefer. See id.
The district court did what we asked. And it stayed within the parameters we set for it. Thus, I dissent.