dissenting:
Although I agree with some of the majority’s reasoning, I must respectfully dissent. I join in the majority’s well-reasoned conclusion that the district court had jurisdiction in this case. As I will explain below, however, I disagree with the majority on the merits of the case. I would reverse and remand for a new trial.
I.
I disagree with the majority’s conclusion that the district court properly excluded the Greenbaum Supplement. ALPA tried to offer the “Greenbaum Supplement” at trial. The original Greenbaum panel opinion stated the following:
The record in this case is voluminous. It contains some 7441 transcript pages, 9 joint exhibits, and thousands of exhibits submitted by both parties. To include pertinent information contained therein in a full Arbitration Opinion and Award, including the Facts and Arguments, will take several months to prepare and document. This, in our view, will unnecessarily delay the issuance of the integrated seniority list. Accordingly, it is the best judgement of this Board that a list should be issued contemporaneously with this Opinion and Award, with full explanation, including facts, arguments and findings to follow as soon as possible.
Shortly after the Greenbaum panel issued its original opinion, however, Continental filed for protection under chapter 11 of the Bankruptcy Code; this filing eventually led to a pilot's strike. The Greenbaum panel decided not to waste its time preparing a detailed opinion, as that opinion may not have had any relevance, depending upon the outcome of the chapter 11 proceedings.
In preparation for this litigation, ALPA asked the Greenbaum panel to issue a supplementary opinion to explain why it had placed plaintiffs in Group 9. The panel issued the Greenbaum Supplement to make this explanation. The district court ruled that the Supplement was inadmissible.1
Plaintiffs stress the procedural irregularities surrounding the preparation of the supplementary opinion, especially the fact that ALPA paid the arbitrators and had the opinion prepared for use in litigation. These arguments, rather than addressing the admissibility of the evidence, concern the weight that the district court should attach to it. Moreover, contrary to the majority’s reasoning, the district court did *618not rely upon these factors in excluding the evidence. Because the district court offered an improper legal basis for excluding the evidence, I believe the court erred in failing to consider the Greenbaum Supplement. Because the Greenbaum panel’s reasons for placing plaintiffs in Group 9 is the critical issue in this appeal, this error was not harmless. A reasonable judge could reach a different conclusion on the merits based upon the contents of the Greenbaum Supplement. I would reverse and remand the case for a new trial.
II.
Even if the district court had not erred by excluding the Greenbaum Supplement, I would still reverse based upon the evidence in the record. Plaintiffs, as the district court found, claim that ALPA’s breach of duty resulted in the Greenbaum panel’s failure to comply with the Morris Award, resulting in their improper placement in Group 9. As plaintiffs frame the issue, “Quite simply, the district court found ALPA breached its duty of fair representation to Plaintiffs when it failed to comply with and enforce the Morris Award in connection with the TXI-CAL Seniority Integration Proceedings. That alone is what this case is about.” Because I find the district court’s findings regarding compliance to be clearly erroneous, I would reverse the decision of the district court even if that court properly excluded the Green-baum Supplement.
To determine whether the Greenbaum panel complied with the Morris Award, one need only look to the Greenbaum Award’s description of plaintiffs. On page 36 of that award, the panel describes Group 9 as follows: “Relatively recent TXI new hires with DOH in 1980 or later and LOS of less than four months as of 11/25/81.” 2 Given this description, one must examine only two pieces of information to determine whether the Greenbaum Award complied with the Morris Award: date of hire and length of service.3
The Morris Award found that the plaintiffs would have had a seniority date of November 19, 1980 — the date on which plaintiffs completed their ground training. According to the TXI seniority agreement, a pilot is not considered to be employed by the company as a pilot until he has completed his ground training. As a result, the pilot’s date of hire and seniority date are identical. The Greenbaum panel used the November 19, 1980, date as plaintiffs’ date of hire as required by the Morris Award.
Turning to the second piece of information, length of service, the Greenbaum panel complied with the Morris Award for this factor, as well. The Morris Award held that the plaintiffs would have been furloughed until July 30, 1981, at which point they would have been recalled to active service and remained active until the critical date that the Greenbaum panel used to determine length of service, November 25, 1981. Under the Morris Award, plaintiffs would have worked 2 days in July, 31 in August, 30 in September, 31 in October, and 25 in November, for a total of 119 days — a little less than four months. The Greenbaum Panel’s description of length of service for Group 9 is therefore consistent with the Morris Award.
The majority admits that plaintiffs would have had no active service in the absence of the Morris Award. Because the Green-baum Award correctly refers to plaintiffs *619as having approximately four months of active service, I find that admission hard to reconcile with the majority’s later conclusion that the two awards are inconsistent. Had the Greenbaum panel chosen to ignore the Morris award, it would not have chosen the four month figure when describing Group 9.
The compatibility of the Greenbaum Panel’s statistics with the Morris Award provides a sufficient reason to reverse the decision of the district court. The majority not only disputes the facial compatibility of the awards, however, but takes the additional step of deciding whether the “Green-baum panel was justified in differentiating between Sims and the Plaintiffs.” The majority appears to substitute its judgment for that of the Greenbaum panel. While giving little consideration to the facial compatibility of the two awards, the majority focuses on where it believes plaintiffs should have been placed in the seniority list. In other words, the majority concentrates on the Greenbaum panel’s use of plaintiff’s employment statistics, rather than on the accuracy of those statistics.
Even if I were to agree that the proper focus was on the qualitative judgments made by the Greenbaum panel, I still would disagree with the majority. The majority focuses on the one year difference in date of hire and concludes that such a small difference could not be significant. Group 2 contains pilots who were hired within six months of pilots in Group 4. Apparently, a small difference in date of hire is significant, when combined with the other factors the panel considered.
Turning to length of active service, the majority concludes that plaintiffs’ four months of service is not significant when compared to Sims’ thirteen months. I disagree. Section 20(E) of the TXI Seniority Agreement provides that a pilot is on probation until he has completed at least twelve months of active service. Although the nine months’ difference in active service between plaintiffs and Sims is a small difference in magnitude, it is a large difference in-kind. The Morris Award recognized the importance of the probationary period, as page 31, paragraph 5 of that award retains plaintiffs’ probationary status until they complete one year of actual, as opposed to constructive, active service.
Accordingly, the Greenbaum Award is consistent with the Morris Award, and the Greenbaum panel had a good reason for placing plaintiffs in Group 9 — their probationary status. Concluding that the majority erred in this regard, I respectfully dissent.
. The district court improperly relied upon our decisions in Iron Workers v. Bowen, 624 F.2d 1255 (5th Cir.1980), and San Antonio Newspaper Guild v. San Antonio Light Div., 481 F.2d 821 (5th Cir.1973). Those cases addressed the question of when the losing party in an arbitration may force a remand to the arbitrator for clarification of the decision; they do not address the question of when a supplementary opinion is admissible. Indeed, in Bowen we held that there was no need to remand an award to the arbitrator because the arbitrator had already clarified the award by writing a supplementary opinion at the request of the prevailing party. This case, then, supports ALPA’s position, as we interpreted the original award based upon a supplementary opinion requested by a party to the arbitration.
I find equally questionable the majority’s and the district court’s conclusion that the award was "clear and unambiguous.” True, Continental had no difficulty implementing the list. Of course, implementing the list only requires reading it. The issue here is the reasoning used in making the list. As the majority does not quote any of the Greenbaum panel’s reasoning in its discussion of the merits, the reasoning is not so “clear and unambiguous.”
. The only further potentially relevant reference to Group 9 appears on page 45 of the opinion, where the panel states, “Finally, we note that notwithstanding this, these CAL furloughees have an earlier DOH and substantial LOS relative to the five TXI Pilots in Group 9, who are fairly recent new hires with little LOS to 11/25/81. Accordingly, they comprise the last group.” This language is obviously consistent with the description of Group 9 quoted in the text.
. The majority provides analysis regarding two other pieces of information considered by the Greenbaum panel — furlough vulnerability and employment status as of November 25, 1981. ALPA’s brief also raises these factors in attempting to justify the Greenbaum Award. I find these factors irrelevant to the issue before us. The majority cannot identify any language in the Greenbaum Award that discusses these factors for Group 9, nor any language which indicates that these factors had anything to do with placing plaintiffs in Group 9.