VII of the Civil Rights Act, constructive discharge occurred where the employee was subjected to what any reasonable senior manager in her position would have viewed as "career-ending action").

    The ALJ found that Complainant failed to prevail under both the "pretext" and "dual motive" analyses commonly applied in discrimination cases. Compare St. Mary's Honor Center v. Hicks, 113 S.Ct. 2742, 2747 (1993); Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 255 (1981); and McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802-803 (1973) (pretext cases) with Price Waterhouse v. Hopkins, 490 U.S. 228, 258 (1989) and Mt. Healthy City Sch. Dist. Bd. of Ed. v. Doyle, 429 U.S. 274, 287 (1977) (dual motive cases). We disagree with the ALJ's use of the pretext analysis. Rather, we employ the dual motive analysis because Complainant has produced " evidence that directly reflects the use of an illegitimate criterion in the challenged decision,' [i.e.,] evidence showing a specific link between an improper motive and the challenged employment decision." Carroll v. United States Department of Labor and Bechtel Power Corp., No. 95-1729, 1996 U.S. App. LEXIS 3813, at *9 (8th Cir. Mar. 5, 1996), quoting Stacks v. Southwestern Bell Yellow Pages, Inc., 996 F.2d 200, 202 (8th Cir. 1993). Evidence of actions or remarks of an employer tending to reflect a discriminatory attitude may constitute direct evidence. Beshears v. Asbill, 930 F.2d 1348, 1354 (8th Cir. 1991). Such evidence does not include stray or random remarks in the workplace, statements by nondecisionmakers or statements by decisionmakers unrelated to the decisional process. Price Waterhouse v. Hopkins, 490 U.S. 228, 277 (1989)(O'Connor, J., concurring). We agree with and adopt the ALJ's findings that the veto of Talbert's transfer to the megawatt improvement program, the alleged refusal to transfer Talbert to Engineering, the initial refusal to run the computer study on Talbert's hypotheticals, the apparent miscommunications, the editing of the May 16, 1991, videotape and the two instances of drug testing were not discriminatory. R. D. and O. at 15-22.

    In dual motive cases under the ERA, a complainant must demonstrate by a preponderance of the evidence that the respondent took adverse action, at least in part, because he engaged in protected activity. If the complainant successfully proves illegal motive, the burden shifts to the respondent to demonstrate by clear and convincing evidence that it would have taken the same action in the absence of the protected activity. 42 U.S.C. § 5851(b)(3)(D); Yule v. Burns International Security Service, Case No. 93-ERA-12, Sec. Dec., May 24, 1995, slip op. at 7-8 and n.7 (courts characterize clear and convincing evidence as more than a preponderance of the evidence but less than evidence meeting the "beyond a reasonable doubt" standard).

       I. Illegal motive

    Complainant Talbert's chain-of-command included the following managers. Rod Webring, the technical services manager, was Talbert's immediate supervisor. Webring reported to Jack Baker, the power plant manager, who in turn reported to Lee Oxsen, the deputy managing

director and chief operating officer. Don Mazur, the managing director and chief executive officer, was superior to Oxsen.

    The ALJ found that Webring, alone, made the decision to transfer Talbert out of the Reactor Engineering Group. R. D. and O. at 10. Baker's testimony suggests broader management involvement in the transfer decision as well as the decision to impose job restrictions. Baker testified: "All along I had decided that there was not a job that I thought was a good match for [Respondent] and [Talbert] within the groups that I controlled, and so that's why we made the management decision that the most likely location for [Talbert] to contribute . . . would be in one of the [other] engineering functions . . . ." T. 249. In addition, Oxsen and Baker were superior in Webring's and Talbert's chain-of-command. It follows that their opinions would influence Webring to some degree. As noted above, the constructive discharge emanated from the transfer and the imposition of job restrictions. Baker admittedly communicated the job restrictions, T. 232-233, and Talbert submitted his resignation on "the day he learned that Mazur left the company and that Oxsen, whom Talbert perceived

   to be his nemesis, took over." R. D. and O. at 5. For these reasons, we consider the motivations of Oxsen, Baker and Webring.

Strict adherence means that you will follow your EOPs, you know, 99 and 1/2 percent of the time. Verbatim compliance says you'll follow them 100 percent. Neither us nor the Commission want to close the door that says . . . you follow EOPs at all costs, because we know that you're talented people, you have lots of experience . . . . We just need to make sure that the threshold in terms of the strict adherence of the procedure, is not down at the 70 percent level, that . . . it has to be an exception . . . in the name of protecting the health and safety of the public. We want you to follow the EOPs and not deviate . . . to protect against commercial risk. . . . I . . . right now, quite frankly, I can't think of any time that we would deviate from our EOPs.

CX 1 (videotape transcription). Talbert responded, "I can think of one." Baker asked: "When's that?" The exhibit 1 transcription continues as follows:

    Talbert: "To trip into region A' with the turbine available ATWS is not only wrong, it is very dangerous . . . ."

    Baker: "Then we need to change that in our EOPs, and exercise that in the form of EOP revision and not exercise that in the form of simulator performance."

    Talbert: "I'm in complete agreement and clearly we will follow the EOPs verbatim. We have a hook in the EOPs currently that is dangerous germane to reactor safety."

    Baker: "But that's an issue that's being pursued by the Owners Group, and the product of that will come out of that."

    Talbert: "That's true."

Baker's testimony is consistent. T. 225-226. Baker also testified that Talbert "did not take the stance that he would deviate from the [ATWS] EOP." T. 235.

    Baker expressed irritation at Talbert's participation, particularly as to its timing. T. 235-237. As the ALJ noted, "Oxsen admittedly resented Talbert's comments." R. D. and O. at 4. Oxsen testified:

From my perspective, Mr. Talbert stood up and offered another example of using his own technical judgment to say that it wasn't always necessary to follow our procedures. From the senior management perspective, after I had spent a considerable amount of time -- and the plant manager, Jack Baker, had reinforced that message -- emphasizing that we had to follow our procedures . . . to have someone stand up then in the aftermath of that and say "not in all cases," I felt I was really, really disappointed that that happened.

The frustration was that I had personally been dealing with an extremely angry regulatory agency that was threatening to wreak havoc on our company and keep it down indefinitely while we beat the operators into submission over this issue of procedure compliance. I had just received feedback from a member of that regulatory body that said, well, hell, one of your people told me today that those procedures are guidelines -- we don't have to follow them, we can use our own technical judgment. On that basis I called the meeting . . . and tried to deliver a message that, by God, this is the way it's going to be . . . . And I was emotional about it because there was so much at stake. So, I think it would be understandable that I would be frustrated after delivering this message, and then someone steps up and says that it doesn't apply in all cases.

T. 204, 216-217. See T. 212. Talbert was not the employee mentioned by Oxsen who caused him to call the meeting. T. 218.

    In examining Respondent's reasons for transferring Talbert out of the Reactor Engineering Group, the ALJ first identified the "controlling question" as being whether the decision was motivated, "in whole or in part, by his protected conduct." R. D. and O. at 8. The ALJ "accept[ed] Talbert's testimony that the May 16 episode was repeatedly mentioned by his superiors when explaining his demotion." Id. In mid-October 1992, Talbert began documenting conversations with various managers directly after they occurred. The resulting journal reflects that during five separate conversations over a two-month period, on November 3, 4 and 19 and on December 12, Baker and Webring cited Talbert's participation in the May 16 meeting as a reason for the transfer and imposition of job restrictions. See CX 3. For example, on the morning of November 3, Baker reiterated that Talbert would not be eligible for any "manager slot" or any jobs that "interfaced" with the NRC, the Institute for Nuclear Power Operators (INPO), the Executive Board, Owners' Groups or Upper Level Management for "fear that during technical interfaces [he] would let something slip that may harm the Supply System's image.'" Talbert's entry continues: "I asked for an example of inappropriate interfacing. [Baker] cited the meeting on the EOPs and my concern on the unstable ATWS." Id. at 4.

    The ALJ also pointed to Webring's testimony "that Talbert's conduct at the May 16 meeting was an example of his poor selection of timing when he brings up issues, how he addresses issues, and how he relates to people,' but was not a major consideration' which led to his termination." R. D. and O. at 8. In particular, when asked whether Talbert's participation in the May 16 meeting was "considered at all" in the employment actions, Webring responded, "I don't believe that it was of any major consideration." T. 261. In a similar vein, Baker testified that Talbert's participation may "have been a factor, it was certainly one of the many data points in [Talbert's] history, but I think the overriding factor clearly rests with the relationship between fuels and the reactor engineering group [a legitimate, nondiscriminatory reason]." T. 229. Baker testified that Webring consulted him about the decision to transfer Talbert out of the Reactor Engineering Group and that he (Baker) concurred. RXJWB-1 at 6.

    The ALJ attempted to distinguish Respondent's various concerns:

The employer contends that on May 16 Talbert raised a proper question, but before the wrong audience, at the wrong time, and in a wrong way. It argues that if Talbert was really concerned about the ATWS problem, he should have asked for an EOPs amendment by filing a so-called PER (Problem Evaluation Report) which every employee had a right to initiate. As Talbert's superiors saw it, Talbert's raising of the ATWS issue on May 16th showed poor judgment because it was obviously at cross purposes with the management's attempt to persuade the crews to follow EOPs.

Id. In this regard, Talbert's journal reflects that "Webring told Talbert that he should have raised the ATWS issue not at the meeting called to demand compliance, but do it at an opportune time, quietly and privately." Id. at 4. Webring made the statement in discussing the reasons for Talbert's transfer. CX 3 at 5. This entry, for the afternoon of November 4, also documents Webring as stating that the May 16 incident "played heavily in [Talbert's] situation" and that Talbert "must have been way out of line." Id. When Talbert objected that he "was being punished for raising a safety concern in [an] open forum," Webring responded that Talbert "shouldn't have raised the safety concern," that he "should have sat mute, said nothing, let it go." Id.

    The ALJ then discussed myriad other reasons cited by Respondent for transferring Talbert out of the Reactor Engineering Group. R. D. and O. at 8-15. The ALJ also identified management's "concern that Talbert, who had no political bone in his body,' was so frank and outspoken . . . and so focused on technical issues . . . that he might press his view of EOPs when dealing with the NRC and thus bring on more troubles for the company." Id. at 14-15. The ALJ then found that "the decision to remove Talbert from the Reactor Group was not related to his May 16 comments, or his later pursuit of the ATWS problem . . . ." Id. at 15.

    Many of the ALJ's previous findings contradict this final finding. They establish that the transfer and job restrictions which caused Talbert to understand that his "nuclear engineering career was over," CX 3, were motivated, at least in some part, by the timing and venue of Talbert's complaint that the ATWS EOP was unsafe. According to Webring, the May 16 meeting exemplified Talbert's "poor timing" in raising issues -- in this instance a safety issue. Baker agreed that Talbert had exhibited a "poor sense of timing." T. 248. Oxsen was angered at the substance of a complaint about the safety of a procedure in a setting where the message was compliance with the procedure. The testimony of Webring and Baker supports a finding that while Talbert's participation in the May 16 meeting may not have been a "major consideration" in Respondent's decision to transfer him and to impose job restrictions, it was a consideration nonetheless.

    An employer may not, with impunity, discipline an employee for failing to follow the chain-of-command, failing to conform to established channels or circumventing a superior, when the employee raises a health or safety issue. Pogue v. United States Dep't of Labor, 940 F.2d 1287, 1290 (9th Cir. 1991) (respondent retaliated because employee bypassed her superior in order to make protected complaint; "chain-of-command" rationale was found to be

pretextual); Ellis Fischel State Cancer Hospital v. Marshall, 629 F.2d 563, 565-566 (8th Cir. 1980), cert. denied, 450 U.S. 1040 (1981) (complainant condemned for "failing to follow normal procedures in bringing problems to the attention of those persons ultimately responsible for the operation of the [defendant and] for making his report to the [NRC] without first formally reporting [the violation] internally"); Leveille v. New York Air National Guard, Case Nos. 94-TSC-3/4, Sec. Dec., Dec. 11, 1995, slip op. at 15-17; Saporito v. Florida Power and Light Company, Case Nos. 89-ERA-7/17, Sec. Ord., Feb. 16, 1995, slip op. at 5-7; Pillow v. Bechtel Construction Company, Case No. 87-ERA-35, Sec. Rem. Dec., Jul. 19, 1993, slip op. at 22-23, appeal docketed, No. 94-5061 (11th Cir. Oct. 13, 1994); McMahan v. California Water Quality Control Board, San Diego Region, Case No. 90-WPC-1, Sec. Dec., Jul. 16, 1993, slip op. at 4-5; Nichols v. Bechtel Construction Co., Case No. 89-ERA-44, Sec. Rem. Dec., Oct. 26, 1992, slip op. at 17, aff'd, 50 F.3d 926 (11th Cir. 1995). Such restrictions on communication would seriously undermine the purpose of whistleblower laws to protect public health and safety. Here, Respondent contends that Talbert should have raised the safety concern through proper channels by filing a Problem Evaluation Report or by bringing it to management's attention "at an opportune time, quietly and privately," rather than by speaking out publicly at a meeting of employees. This rationale does not support discipline for protected activity under the ERA.

    We also find that the manner in which Talbert raised the safety concern was not so "disruptive" as to be "indefensible under the circumstances," and, accordingly, that Talbert did not lose protection under the ERA. See Martin v. The Department of the Army, Case No. 93-SDW-1, Sec. Rem. Ord., Jul. 13, 1995, slip op. at 5 (standard for addressing behavior associated with exercise of whistleblower rights). Talbert raised the concern during a "question and answer" period in which Respondent elicited audience participation. The concern was germane to the issue under discussion. Oxsen had advocated verbatim (100%) compliance with all EOPs, and Baker had stated that he could not "think of any time that [they] would deviate from [their] EOPs." CX 1. Talbert simply responded that he could "think of one" -- a specific instance where compliance could compromise safety. Talbert was not abrasive, he did not incite any disobedience on the part of others, nor did he "take the stance that he would deviate from the EOP." T. 235 (Baker). When Baker responded that the issue was being addressed in another forum, Talbert agreed without argument and did not pursue the concern further at the meeting. We find nothing objectively disruptive about Talbert's participation. That Oxsen perceived it as an attempt to subvert his message, T. 204, 216-217, serves rather as a measure of his frustration at being unable to satisfy the NRC in the matter of operator recertification.

    Complainant thus proved that Respondent's adverse action was motivated at least in part by an impermissible criterion, namely his participation in the May 16, 1991, meeting.

       II. Legitimate motivation

    The ALJ found that Respondent demonstrated by clear and convincing evidence that it would have taken the same action even if Complainant had not engaged in protected activity. We agree with regard to the decision to transfer Complainant out of the Reactor Engineering Group. Accordingly, Respondent avoids liability for this adverse action because it would have taken the same action absent illegal motivation.

the Engineering Department. Powers considered the work to be "meaningful and important," and Talbert expressed interest. T. 276-277. Powers testified:

At the same time that I was preparing to talk with Mr. Talbert about opportunities in my organization, Mr. Baker and Mr. Webring were scheduled to meet with him regarding his accountability for the power oscillations event. Their meeting occurred before I could speak with Mr. Talbert and when I met with him thereafter . . . he was very upset and fairly difficult to talk with. . . . The first job that I outlined was developing a staff to support the new WNP-2 simulator. I then went on to describe a second position . . . which was needed to adopt our Probablistic Risk Assessment (PRA) capabilities to an in-line operations decision-support function and to generally assist in the completion of the PRA program. Finally, I discussed a third . . . position . . . as a fuel cycle analyst. . . . Mr. Talbert was excited about the PRA and simulator positions, but was less enthused about the fuel management position.

RX CMP-1 at 3-4. Talbert's journal reflects that this conversation occurred on November 2. Talbert also documented discussions with Baker and Webring about job limitations which occurred on November 3. Ensuing journal entries show a deteriorating relationship with these two managers. CX 3 at 6-9.

    In so finding, we recognize that Talbert's journal portrays Baker as being adamant that the job limitations would be imposed and Webring as being "tight lipped" about who in the chain-of-command had mandated the limitations. Eventually, in the November 3 discussion, Baker revealed that "Upper Management" would not allow Talbert to take two of the positions offered by Powers. When Talbert raised the possibility that Baker's job limitations might apply in the Engineering Department, however, Powers responded that he was unaware of any such limitations. Powers testified: "I told Bob that I was running an organization irrespective of whether limitations were or were not created, he was really going to answerable and accountable to me, and I would lay out the ground rules on which I would judge success or failure." T. 278-279. Nevertheless, Powers "committed to" Talbert that he would clarify with his peers and superiors "whether there was any concern on anybody's part involved in Supply System management . . . as to whether there were any limitations." T. 279. Powers further represented that he would get back to Talbert with a response. Given that Powers and Talbert were engaged in employment negotiations, we believe that Talbert's unwillingness to await Powers's response further supports the finding that his resignation was voluntary.

    While the decision to transfer Complainant was motivated in part by protected activity, Respondent has demonstrated by clear and convincing evidence that it would have taken the same action in the absence of the protected activity. In addition, while certain of Respondent's managers imposed job limitations in part because of the protected activity, the limitations were not universal and did not constitute "career-ending action." Complainant was not constructively discharged when he resigned his employment. Accordingly, the complaint IS DISMISSED.

    SO ORDERED.

      DAVID A. O'BRIEN
      Chair

      KARL J. SANDSTROM
      Member

      JOYCE D. MILLER
      Alternate Member