Opinion of the Court
DARDEN, Judge:On February 13, 1970, a general court-martial found the appellant guilty of four specifications alleging larceny in violation of Article 121, Uniform Code of Military Justice, 10 USC § 921. The court sentenced him to a bad-conduct discharge, total forfeitures, confinement at hard labor for nine months, and reduction to the lowest enlisted grade. The convening authority approved the findings and sentence, but he suspended execution of the bad-conduct discharge and included a provision for automatic remission.
After a majority of a panel of the United States Army Court of Military Review proposed to set aside the findings and sentence and to dismiss the charge, a judge of another panel requested reconsideration en banc of the proposed decision. A majority of the judges of the Court of Military Review voted to reconsider the case en banc. After such reconsideration, an “OPINION OF THE COURT” afiirmed the findings and sentence. Contending that the en banc opinion interfered with the panel’s statutory duty to decide his case, the appellant by motion requested the *596panel to enter its own decision. The panel held it had no jurisdiction to entertain the motion. Following the appellant’s petition for a writ of mandamus, this Court treated the petition as one for a grant of review and granted it.
The real issue is whether the procedure this case followed in the Army-Court of Military Review conflicts with this Court’s decision in United States v Chilcote, 20 USCMA 283, 43 CMR 123 (1971). We hold that as in Chilcote the procedure does conflict with the statute that authorizes the Courts of Military Review to sit in panels or en banc.
In Chilcote we held that a decision by a panel of the United States Navy Court of Military Review could not be reconsidered and reversed en banc. We recognized the desirability of en banc consideration in achieving a uniformity among panels of the same court. We decided ' Chilcote, however, against a background of congressional opposition to the reversal of a panel decision favorable to an accused by another panel of the same court. See Chilcote, supra, at page 285.
Under the Standing Operating Procedure of the United States Army Judiciary, on cases in which the panels prepare opinions, a draft of the proposed opinion is circulated among all the judges of the whole court. If one of the judges suggests en banc consideration, and if a majority of the whole court agrees, the panel’s proposed decision and proposed opinion are withdrawn. In the instant case, no pleadings were filed with the en banc court and argument before the en banc court was neither requested nor waived.
The first part of the Government’s argument is directed against what the Court decided in Chilcote. In developing this argument the Government points to a Sectional Analysis in the Senate Report1 accompanying the bill that became the Military Justice Act of 1968.2 This Sectional Analysis referred to an objective of providing for “sound internal administration” within the Courts of Military Review. This reference to sound internal administration the Government views as evidencing an intent to eliminate the inconsistent and contradictory opinions by the separate and autonomous boards of review that had functioned under the 1950 Act.3 But the Sectional Analysis used in the Report was furnished by the Department of Defense when it proposed the language of what is now Article 66(a) of the Code.4 If providing for sound administration was intended to mean that the full court could reverse a panel decision, the chosen language was uninformative, if not cryptic. Since the Senate Report did not enlarge on the Department’s explanation of this objective, that explanation we consider not persuasive enough to alter the Chil-cote decision.
Next the Government points to the statutory basis for en banc hearings in the United States Courts of Appeals and suggests that Congress used the appellate courts established under Article III of the Constitution as a model when it abolished boards of review and authorized the creation of Courts of Military Review. But the statutory language governing en banc consideration by United States Courts of Appeals is so much more explicit on this subject than is Article 66(a) that the absence of equal specificity in Article 66(a) could be cited to support a result different from the one the Government urges. The two positions are juxtaposed below for comparison:
*597“Cases and controversies shall be heard and determined by a court or division of not more than three judges, unless a hearing or rehearing before the court in banc is ordered by a majority of the circuit judges of the circuit who are in regular active service. A court in banc shall consist of all circuit judges in regular active service. A circuit judge of the circuit who has retired from regular active service shall also be competent to sit as a judge of the court in banc in the rehearing of a case or controversy if he sat in the court or division at the original hearing thereof.” [28 USC § 46(c).]
“Each Judge Advocate General shall establish a Court of Military Review which shall be composed of one or more panels, and each such panel shall be composed of not less than three appellate military judges. For the purpose of reviewing court-martial eases, the court may sit in panels or as a whole in accordance with rules prescribed under subsection (f). Appellate military judges who are assigned to a Court of Military Review may be commissioned officers or civilians, each of whom must be a member of a bar of a Federal court or of the highest court of a State. The Judge Advocate General shall designate as chief judge one of the appellate military judges of the Court of Military Review established by him. The chief judge shall determine on which panels of the court the appellate judges assigned to the court will serve and which military judge assigned to the court will act as the senior judge on each panel.” [Article 66(a), Uniform Code of Military Justice, 10 USC § 866(a).]
*598Our decisions in Mercer v Dillon, 19 USCMA 264, 41 CMR 264 (1970), and Enzor v United States, 20 USCMA 257, 43 CMR 97 (1971), the Government cites in support of a conclusion that a draft opinion is not a decision until it is signed, dated, and filed with the clerk. Those opinions, however, do not address themselves to the precise issue with which we are concerned here.
For the purpose of this decision we disregard the statistics tending to show that en banc consideration has not resulted in decisions that are unsound or unjust. We entertain no thought that panel deliberations are more sound or more just than those of the whole court. We tried to forestall the drawing of any such inference from our opinion in CMlcote. But this is an issue we must decide by statutory construction, not by our evaluation of results under one possible construction of the statute.
Under the Navy procedure that was the subject of CMlcote, the panel decision was published. If we held that the Army procedure in issue is consistent with Article 66(a), reconsideration of Navy panel decisions could be achieved simply by adoption of the Army practice. A difference in result ought to be founded on something more substantial than whether the panel decision was publicized before the sense of the full body was sought. In the practical consequences to an accused, the procedure followed by the Army Court of Military Review is no less a reconsideration than if the Navy practice were in effect.
As in CMlcote, we construe Article 66(a) as meaning that cases are to be reviewed and decided by panels of the court or by the entire court, but that the same case may not be decided by two different groups of judges within a Court of Military Review. Before a panel has made a determination on the merits, a case referred to it may be referred for en banc consideration without conflicting with Article 66(a). But except for such a contingency, when a panel of a Court of Military Review has had a case referred to it, that panel normally is obligated to decide the case and to make its decision public instead of informing preliminarily only the other members of the court.
We continue to be aware that en banc reconsideration has obvious advantages, not the least of which is that it would obviate this Court’s consideration of some certified issues arising from panel decisions. But again this is not the test. Unless pleadings, briefs, and arguments are directed to the group of judges that decides the case, some of the normal attributes of appellate review are missing.
Our decision is that the Army arrangement conflicts with the Article 66(a) provision that review is to be by a panel or by the whole court. Accordingly, the en banc decision of the United States Army Court of Military Review is reversed. The record is returned to the Judge Advocate General of the Army for action not inconsistent with this opinion.
Chief Judge Quinn and Senior Judge FeRguson concur.Sectional Analysis, section 2(27), Senate Report No. 1601, 90th Congress, Second Session (1968).
Public Law 90-632, 82 Stat 1335.
Act of May 5, 1950, Public Law 506, 64 Stat 107.
See letter of July 1, 1968, from the Secretary of the Army to the Chairman of the Senate Committee on Armed Services in response to a request of that Committee for the views of the Department of Defense on S. 2009, 90th Congress, on file in Committee on Armed Services, United States Senate.
*597Another Government contention construes our opinion in Chilcote as prohibiting only an en banc reconsideration of a panel determination of fact favorable to an accused. Fact-finding power by the panels appears to have been the primary source of congressional concern over reversal by a different board of review of a decision by another board. See Hearings before a Subcommittee of the Committee on Armed Services, House of Representatives, on HR 2498, 81st Congress, First Session (1949), at pages 1191-1195, 1201-1207. Separation of factual and legal determinations by panels was an available alternative in Chilcote, but it is one this Court did not adopt. As appellate defense counsel note, the line of demarcation is not always clear and many decisions turn on mixed questions of fact and law. For reasons developed in Chilcote we thought the statutory language and history of the Military Justice Act of 1968 were not enough to establish a congressional intent to reverse its earlier position against reconsideration of panel decisions in favor of an accused. We indicated our awareness of the desirability of en banc consideration to avoid conflicting holdings by different panels of the same court, but we thought the function of weighing such desirability against possible disadvantages of en banc consideration to an accused is one properly for Congress.
The final Government argument suggests that Chilcote disapproved en banc reconsideration of panel decisions and that under Army practice the draft of a panel determination that is circulated among members of the whole court is not a decision in any sense of the word. According to this argument, if the case is to be considered en banc the remaining copies of the draft panel opinion are destroyed. If the case is not to be considered en banc, the draft becomes an opinion only when the signatures of the panel judges are affixed to the original copy. Hence no panel decisions exist, but only drafts of proposed opinions that, in the absence of en banc consideration, become the decision of the entire court.