concurring in the judgment:
I concur in the majority’s remand in light of United States v. Ameline, 409 F.3d 1073 (9th Cir.2005) (en banc). I write separately, however, to disagree with the majority’s acceptance of the use of a sentencing council in determining Brigham’s sentence. Brigham entered into a plea agreement with the Government for a sen*671tence of twenty-four months. After meeting with a sentencing council, the District Judge increased Brigham’s sentence to thirty-seven months. Neither Brigham nor the public will ever know what impact the council had on Brigham’s increased sentence because neither party in the case was allowed to attend the sentencing council meeting, and a record of the exchange was not disclosed. Brigham’s mere knowledge that a sentencing council was used in his case does not remedy this harm.
The majority underestimates the “potential for abuse,” the “doubtful appearance, and the possible misunderstanding” of sentencing council proceedings that take place outside the defendant’s presence. United States v. Spudic, 795 F.2d 1334, 1344 (7th Cir.1986). In Spudic, for example, the Seventh Circuit held that, while a judge may confer with his particular probation officer, the use of a sentencing council made up of several probation officers could not be sanctioned regardless of its supposed benefits. Id. Even though Spudic dealt with a sentencing council made up of probation officers, and this case deals with a sentencing council consisting of judges, the same principle prohibiting ex parte communications between the sentencing judge and outside parties applies to either form of sentencing council. See id. at 1343-44 (describing probation officer sentencing council as an ex parte conference). Canon 3 of the Judicial Code of Conduct, which dictates protocol for federal judges, explicitly requires that a bench officer “neither initiate [n]or consider ex parte or other communications concerning a pending or impending proceeding.” See Fed. R. Decisions, Canon 3(4). When a judge relies on secret discussions with other judges to make a decision in his case, the judge necessarily initiates ex parte communications.
In this case, the Government contends that Brigham was not harmed by these ex parte communications because the District Judge only discussed case law that was favorable to the defendant with the council. Even if that is true, “we are concerned with the institutionalized use of this sentencing council procedure because of the concern and doubts which it can understandably foster in the minds of defendants, their counsel, and the public.” Spudic, 795 F.2d at 1343. Because the council meets ex parte, the full extent of any possible harm caused by the sentencing council will never be known by either party unless the District Court chooses to disclose it. Faith in the integrity of sentencing judges leads me to agree that a judge should normally not need to defend or explain every element he relies on in his sentencing decision. See id. at 1344. Such integrity, however, must be bolstered by transparent procedural safeguards, such as public hearings that do not require the parties and the public “to accept the sentencing council deliberations on faith.” Id. Secret sentencing councils, rather than being justified by a judge’s integrity, may call it into question. This is the harm caused by ex parte communications.
The majority notes that United States v. Gonzales, 765 F.2d 1393, 1398-99 (9th Cir.1985), does not disapprove of Oregon’s sentencing councils. However, Gonzales also does not expressly approve of such councils; this Court’s analysis was directed at whether communications between the sentencing judge and the probation officer during the sentencing process were improper. Id. (rejecting defendant’s claim that probation officer acted improperly by engaging in “ex parte advocacy”). This Court previously established in United States v. Davis, 527 F.2d 1110, 1112 (9th Cir.1975), cert. denied, 425 U.S. 953, 96 S.Ct. 1729, 48 L.Ed.2d 196 (1976), that a judge can discuss the presentence report *672and sentence with the probation officer outside the defendant’s presence. Gonzales adopted the Davis holding. 765 F.2d at 1398. This Court has not yet directed its analysis at sentencing councils, such as Brigham’s, that are comprised exclusively of judges. Yet, the reasoning of Spudic demonstrates use of such a council, consisting of either probation officers or judges, is an error.
In addition to constituting a troubling ex parte communication, the use of a sentencing council erodes the well-established principle that federal judges should be independent and insulated from group pressures. Article III of the Constitution provides life tenure and undiminished due compensation to federal judges to preserve their autonomy.
Indeed, early constitutional debates in this country underscore the importance of judicial independence and insulation:
[The] independence of ... judges is equally requisite to guard the Constitution and the rights of individuals from the effects of those ill humors, which ... the influence of particular conjunctures ... sometimes disseminate among the people themselves, and ... have a tendency, in the meantime, to occasion dangerous innovations in the government, and serious oppressions of the minor party in the community.
The Federalist No. 78 (Alexander Hamilton). Sentencing councils require judges to be privately inter-dependent, which makes the judicial process dangerously susceptible to improper communal pressure.
A judge, of course, is not prohibited from consulting other judges. See Fed. R. Decisions, Canon 3(4) Commentary. But when proceedings are carried out in secret, it is not possible to determine whether a decision was the result of permissible consultation or impermissible pressure. One of the original purposes of sentencing councils was to reduce sentencing disparity by placing “group pressure [on judges] to conform.” Federal Judicial Center, The Effects of Sentencing Councils on Sentencing Disparity 1 (1981). In Spudic, the Seventh Circuit described this potential for group pressure and misunderstanding:
The sentencing council may have an unrecognized influence on the sentencing judge causing the judge to abide by the council consensus. That could lead to the further concern that the impact of what is subsequently presented in open court at sentencing will be minimized, that the sentence will be largely foreordained, and that the judge therefore enters the actual sentencing hearing without an open mind.
795 F.2d at 1343 (emphasis added). Because the sentencing council met in secret and may have had “an unrecognized influence on the sentencing judge,” it is not possible to know if the increase in Brigham’s sentence was the result of impermissible group pressure. Id. We lack basic proof — a record.
The fact that the sentencing council in Spudic consisted of probation officers and Brigham’s council consisted of judges makes little difference to this analysis. The peer pressure exerted by fellow judges, with whom the sentencing judge must interact on a regular basis both formally and informally, is likely to have a greater impact than any concern for the staff morale of probation officers. Judges are also just as likely to share “additional pertinent adverse information about the defendant” with the sentencing judge as are probation officers who did not participate in the preparation of the presentence report. See id. Therefore, a sentencing council consisting of judges, as opposed to probation officers, is as much if not more damaging to the interest of the defendant *673in being sentenced in an individualized, transparent manner.
Apart from the injustice that results from ex parte communications, sentencing councils should be abolished because they are unnecessary. To the extent that the Sentencing Guidelines still play a role in a judge’s determination of a sentence, sentencing councils frustrate any transparency provided by the advisory use of the Sentencing Guidelines. United States v. Booker, 543 U.S. 220, 259, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005) (finding that “[without the ‘mandatory’ provision, the[Sentencing Reform Act] nonetheless requires judges to take account of the Guidelines together with other sentencing goals”). The Sentencing Reform Act of 1984 and the United States Sentencing Commission both operate to reduce sentencing disparities, supplanting the original purpose of sentencing councils. S. Rep. No. 98-225, at 52 (1983), reprinted in U.S.C.C.A.N. 3182, 3236 (explaining how a primary goal of sentencing reform is the elimination of unwarranted sentencing disparity). In fact, when Congress created the Sentencing Guidelines, only presen-tence reports, not sentencing councils, were meant to supplement the Guidelines. Id. at 53. Congress’s ultimate goal was to allow “each participant in the system ... [knowledge as to] what purpose [was being] achieved by the sentence in each ... case.” Id. at 59. Each time a secret sentencing council convenes to assist a judge in rendering a decision, this goal is frustrated.
The use of sentencing councils is even more troubling in light of the new, advisory nature of the Sentencing Guidelines. Booker, 543 U.S. at 226, 125 S.Ct. 738. Judges now have greater discretion to impose sentences based on specific facts not enumerated in the Guidelines. What now protects a defendant from having outside, undisclosed information influence the deliberations of the council and, subsequently, the judge’s discretionary sentence?
Sentencing councils promulgate decisions that are susceptible to impermissible group pressure. An informed judge, when conducting sentencing, need only use the Sentencing Guidelines, the presentence report, and arguments and evidence produced in open court to arrive at a decision grounded in law. We should no longer tolerate these secret ex parte proceedings.