concurring in part and dissenting in part:
I agree with the panel’s disposition of McCoy’s claim for reconsideration of the *1065restitution element of her sentence. I disagree, however, with its conclusion that McCoy has waived her objection to the two-point obstruction bump for the offense level on the perjury charge.
Sentencing nowadays proceeds under the United States Sentencing Guidelines and involves complexities and contingencies not unlike those of the Internal Revenue Code. Just as under the IRC an increase in income at certain levels will reduce the benefit of a taxpayer’s deductions, see 26 U.S.C. § 68 (reducing allowable deductions by 3% of the excess of adjusted gross income over applicable statutory threshold); see also id. § 151(d)(3) (reducing personal exemption amount by 2% for every $2,500 in income over applicable statutory threshold), so under the Guidelines a decrease in the “offense level” for the higher-scored of two related crimes may end up changing the effect of the lower-scored crime, and more particularly, change the impact of the offense level assigned to the lesser crime. As a result, an issue that at one point appears purely hypothetical may move into prominence once adjustment is made elsewhere in the calculation.
This of course is what happened to McCoy. See Maj. Op. at 1060-61. On her initial sentencing the district court found her offense level for the false statement charge to be 20 and for perjury 14. Because the difference of six (20 minus 14) fell between five and eight points, under § 3Cl.l’s provision for a “multi-group adjustment” the perjury charge caused a one-point upward bump in her combined score, which the district court accordingly set at 21. If McCoy had at that point argued for a reduction of her perjury offense level from 14 to 12, as she now does, she would rightly have been told it was immaterial; the multi-group adjustment is the same for an eight-point differential as for a six-point differential, and the perjury conviction’s only impact on the combined offense level was via that “adjustment.”
On her first appeal we saw merit in her claim that the district court had perhaps given her an improper two-point bump for her managerial role in the false-statement offense. Unsure as to the district court’s understanding of the standard and as to its view of the facts, we remanded. United States v. McCoy, 242 F.3d 399, 409-10 (D.C.Cir.2001). On remand, the district court in fact deleted the two-point “managerial” bump. With the false statement charge falling to an offense level of 18, the gap between it and the perjury score fell to four (18 minus 14). Following § 3C1.1, the district raised the multi-group adjustment to two points, now producing a combined score of 20. With the reduction in the false statement offense level, the issue McCoy now raises became material: reducing the perjury offense level from 14 to 12 would restore the gap to six points and thus take the multigroup adjustment back down to one.
Accordingly, McCoy then put before the district court her complaint that it had erroneously given her perjury charge a two-point bump for “obstruction of justice.” The district court refused to entertain the claim because, in its view, the only issue properly before it was the two-point “managerial” bump in the false statement charge. Resentencing Hearing Tr. at 13 (May 9, 2001). (Not exactly: it assumed that modifying the multi-group adjustment was in order.) The majority today holds that McCoy’s argument against the obstruction supplement in the perjury context is not “newly relevant” and that McCoy waived it by failing to raise it during the initial hearing and appeal. This seems to me both an unsound reading of our decision in United States v. Whren, 111 F.3d 956 (D.C.Cir.1997), and likely to *1066lead to inefficient and needlessly complex sentencing challenges.
In Whren we held that
... upon a resentencing occasioned by a remand, unless the court of appeals expressly directs otherwise, the district court may consider only such new arguments or new facts as are made newly relevant by the court of appeals’ decision — whether by the reasoning or by the result.
Id. at 960 (emphasis added). We also said, “A defendant should not be held to have waived an issue if he did not have a reason to raise it at his original sentencing.” Id.
The present case exposes an ambiguity in Whren. As I understand the majority, an initial appeal renders an issue “newly relevant” if but only if the decision opens up an issue that previously could not have been even potentially relevant. Apart from issues injected by the court of appeals on its own hook, and presumptively from out of left field (which we’re generally not supposed to do), I’m not sure what issues could fall into this group — apart, obviously, from any “newly relevant” issues that the court of appeals explicitly tells the district court to consider on remand.
The more natural reading, I think, would be that an issue is “newly relevant” if it will, entirely as a result of changes resulting from the first appeal, for the first time have an actual impact on the sentence. The only circuit court to adopt or even seriously consider Whren, the First Circuit in United States v. Ticchiarelli, 171 F.3d 24, 32-33 (1st Cir.1999), so understands it. In the initial sentencing there, the district court had fitted a large weight of “hashish oil” into the guidelines system by treating each kilo of hashish oil as the equivalent of 50 kilos of marijuana. The court of appeals reversed, saying that the oil must be treated as equal to marijuana. On remand, the defendant then raised an issue about computation of the weight of the drugs' containers, an issue that had been — because of the numbers in question — irrelevant under the district court’s initial (erroneous) treatment of hashish oil. The district court refused to hear the claim, treating its omission from the first appeal as a waiver. The court of appeals rejected the waiver theory, quoting Whren at some length and reading it to allow the defendant to raise his “container” issue in the second appeal, because he had not had “sufficient incentive to raise the issue in the prior proceedings.” Id. at 33. Of course in this circuit, under the panel’s view, fear of this court’s waiver doctrine would supply an incentive, but the First Circuit plainly saw incentive as depending on whether the issue could change the sentence under the view taken by the district court on the other issues. As no change was possible under those views, there was no waiver. This strikes me as the sound and natural reading of Whren.
Besides draining the Whren exception of virtually all meaning, the panel view here imposes an undue burden on counsel and will ultimately increase the burdens on the court. The impact on defense counsel is plain. Counsel must, in the initial appeal (and therefore, of course, at the initial district court sentencing), consider every step of the presentence report’s reasoning, to determine whether any such step might not conceivably have an adverse effect down the road if the defendant won on any of its proposed changes — no matter how immaterial the step may be on the view actually taken by the report. And, of course, he or she must articulate and argue these seemingly moot issues before the district court and the court of appeals.
The court may suppose that the resulting burden will fall only on counsel (though I do not know why a court should disre*1067gard a waste of others’ resources). But even from a narrowly judicial perspective, the rule will generate needless complexity, causing briefs to be cluttered with second-order issues that are hypothetical and in many instances ultimately irrelevant (in fact, as we shall see below, in most instances). Cf. Crocker v. Piedmont Aviation, Inc., 49 F.3d 735, 740 (D.C.Cir.1995) (holding that appellees should not be forced “to put forth every conceivable alternative ground for affirmance,” and expressing serious concerns about increasing “complexity and [the] scope of appeals”); cf. also Field v. Mans, 157 F.3d 35, 41-42 (1st Cir.1998); Exxon Chemical Patents, Inc. v. Lubrizol Corp., 137 F.3d 1475, 1478-79, 1482 (Fed.Cir.1998); Laitram Corp. v. NEC Corp., 115 F.3d 947, 954 (Fed.Cir.1997). It might be consoling to think that counsel will provide the court with a handy roadmap: for example, “If the court agrees with point 2, then points 3 and 4 are irrelevant, and the court should proceed directly to point 5.” Nothing in my experience on the bench supports the notion that counsel will do so; rarely if ever do counsel even bother to say, for example, “For reversal, appellant must prevail on both arguments I and II.”
While a number of circuits have adopted a rule of de novo resentencing, see, e.g., United States v. Jennings, 83 F.3d 145, 151 (6th Cir.1996); United States v. Atehortva, 69 F.3d 679, 685 (2d Cir.1995); United States v. Cornelius, 968 F.2d 703, 705 (8th Cir.1992), there is good reason not to go so far. We said in Whren that such a rule “is in essence a license for the parties to introduce issues ... that they should have introduced at the original sentencing hearing.” 111 F.3d at 959. Quite true. A better solution is to give Whren its most obvious reading, as the First Circuit did in Ticchiarelli: allow counsel to defer any second-order issue, i.e., any issue that can come into play only if the defendant should secure victory on a first-order issue. See also United States v. Parker, 101 F.3d 527, 528 (7th Cir.1996) (“Only an issue arising out of the correction of the sentence ordered by this court [can] be raised in a subsequent appeal.”).
Reversal rates may influence one’s choice among the alternative readings of Whren. Multiple appeals from a single case cost judicial (and lawyers’) resources, but increments in the complexity of single appeals also cost resources. If district courts are seldom reversed, the broad (First Circuit) reading of Whren makes sense. After all, if a litigant only has a 10% chance of reversal on the first-order issue, there is little reason to burden courts with second-order issues that will become relevant only 10% of the time. In contrast, if district courts are frequently reversed on first-order issues, then there is more to be said for the narrow reading of Whren, as the frequency of second-order issues becoming relevant will concomitantly increase.
In fact it appears that reversal rates are low. According to statistics kept by the Administrative Office for the United States Courts, the reversal rate in this circuit is 12.9% generally, and 12.1% in criminal cases. For all circuits combined, the reversal rate is 9.5% generally, and only 6.3% in criminal cases. See Judicial Business of the U.S. Courts 102 tbl. B-5 (2000) (reporting rates for appeals terminated on the merits for the 12-month period ending September 2000).1 Reversal rates for sen*1068tencing issues, of course, may be higher, but given the frequency of sentencing issues, a high reversal rate on such issues would be reflected in high average rates. Thus the data tend to support the broader reading of Whren — allowing a defendant to raise on his or her second appeal issues that were entirely hypothetical on the first round.
I respectfully dissent from the court’s disposition of McCoy’s perjury charge claim. As the majority finds that claim waived, I do not reach the merits of the issue. Cf. Daingerfield Island Protective Society v. Babbitt, 40 F.3d 442, 448 (D.C.Cir.1994) (Wald, J., dissenting in part); cf. also Boggs v. Rubin, 161 F.3d 37, 44 (D.C.Cir.1998) (Rogers, J., concurring in part and dissenting in part).
. The reversal rates reported here are those shown in the Administrative Office's report. The statistics may be somewhat imprecise for purposes of this analysis because cases need not be affirmed or reversed, but can be disposed of in other ways. Nonetheless, recalculating the rates to consider only cases that were either affirmed or reversed yields simi*1068lar numbers: for this circuit, 13.2% generally and 14.6% in criminal cases; for all circuits, 9.9% generally and 7.1% in criminal cases. Statistics for earlier years yield similar results. See Judicial Business of the U.S. Courts 108 tbl. B-15 (1999); Judicial Business of the U.S. Courts 114 tbl. B-5 (1998).