Permanente Medical Group

PERMANENTE MEDICAL GROUP 1143 The Permanente Medical Group, Inc. and Engineers And Scientists Of California Local 20, IFPTE, AFL–CIO, CLC and Kaiser Foundation Hospi- tals and the Permanente Medical Group, Inc. and California Nurses Association. Case 32– CA–15032 and 32–CA–15084 October 31, 2000 DECISION AND ORDER BY CHAIRMAN TRUESDALE AND MEMBERS FOX AND HURTGEN On December 19, 1996, Administrative Law Judge Mary Miller Cracraft issued the attached decision. The General Counsel filed exceptions and a supporting brief, the Respondent filed an answering brief to the General Counsel’s exceptions, and the General Counsel filed a reply brief. The National Labor Relations Board has delegated its authority in this proceeding to a three-member panel. The Board has considered the decision and the record in light of the exceptions and briefs and has decided to affirm the judge’s rulings, findings, and conclusions and to adopt the recommended Order. I. FACTS The Respondent provides health care services to its members at various hospitals and medical clinics throughout northern California. There are two charging party unions in this case. California Nursing Association (CNA) represents approximately 7500 of the Respon- dent’s registered nurses, and Engineers and Scientists of California Local 20, IFPTE, AFL–CIO, CLC (ESC) represents approximately 680 medical technologists. In addition, although not a Charging Party, Local 250 of the Hospital & Health Care Workers Union represents ap- proximately 13,000 employees in various technical ser- vice and maintenance classifications. The Respondent has an ongoing bargaining relationship with each of the unions. In 1993 the Respondent, working with Anderson Con- sulting, began developing a project known as the “Gate- ways Project.” The goal of the project was to incorpo- rate the concept of “patient-focused care” on an experi- mental basis into several small new hospitals that were then under construction. The principal of patient-focused care was to improve the method by which the Respon- dent delivered health care services to its patients. In February 1995 after bargaining with its unions, the Respondent implemented the Gateways Project at a new hospital in Fresno, California. That implementation is not a part of this case. In response to significant changes in the health care industry, and in light of a nongrowth in membership, the Respondent considered building on its “patient-focused care” concept and applying it in a larger setting. In early 1995, again working with Anderson, it developed a con- cept called “member focused care” (MFC) which sought to increase patient and family involvement in care and to reorganize care management. At this time, the Respon- dent contacted representatives of its largest unions, CNA and Local 250, and informed them of the process that it intended to follow before implementing MFC. This process would include focus groups in July 1995 to ex- plain the concept and discuss how the approach might be structured. These focus groups would be followed by design team meetings in August which would develop recommendations to be submitted to management. Man- agement would then prepare a proposal to present to the unions for bargaining. Focus groups were held from July 13–21. Representa- tives from ESC, CNA, and Local 250 attended these ses- sions. Of the approximate 150 participants there were four employees represented by ESC and eight repre- sented by CNA. Participation was voluntary. The pur- pose of the meetings was to provide an overview of the Gateways model and the proposed MFC model. Addi- tionally, the focus groups were to obtain feedback from participants about the feasibility of the model and its implications for equipment, facilities, and training. Prior to the meetings, the Respondent’s labor relations staff briefed the project leaders on the need to ensure that none of the focus group sessions involved “bargaining” or discussions about bargainable subjects. Discussions included an analysis of current licensed jobs and of pro- posed new nonlicensed job categories. Some employees described the sessions as brainstorming. The complaint includes no allegations that the Respondent violated the Act during this phase of the MFC project. After the focus group sessions CNA representatives notified the Respondent that they believed that the real purpose of MFC was to assign nurses’ duties to other nonlicensed personnel, thereby transferring work out of the unit. They demanded immediate bargaining. In re- sponse, the Respondent advised that it recognized its duty to bargain, that it had not yet decided on a final pro- posal, and that it would bargain after it had formulated such a proposal. ESC representatives expressed similar concerns. The Respondent responded that it was seeking employees’ involvement to hear their ideas and concerns, that noth- ing was final, and that they would bargain as required. In response to this reassurance ESC indicated that it would continue to participate in the next phase—the de- sign phase. 332 NLRB No. 106 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1144 Prior to beginning the “design phase,” the Respondent issued a statement to all participating employees. It ad- vised them that their input was needed because of their professional expertise, that their participation was volun- tary, and that involvement in the design phase did not entail bargaining or setting working conditions. The Respondent reasserted its commitment to bargain with the Unions after it had developed a suitable model for proposal to the Unions. Prior to the meetings, the Re- spondent’s labor relations consultant met twice with the project leaders to go over ground rules on how to ensure that the meetings did not involve any bargaining over wages, hours, or other mandatory subjects. Of the 180 participants, in the design phase, about 51 were bargaining unit employees. This included 27 mem- bers of Local 250, 21 members of CNA, 1 ESC member, and others. (Although ESC withdrew its authorization for its employees to participate in this phase, one of its employees attended independently.) The design meetings began on October 6 and ended around December 21. The employee participants were divided into five teams. Each team met approximately six–eight times over the 2-1/2 month period. Three of the teams dealt with job-related issues. The members of these teams discussed in detail the tasks currently being performed by employees in the various job classifica- tions to the extent that the tasks directly affected patient care delivery. Each team was asked to come up with ideas or recommendations for later presentation to man- agement on how the functions assigned to that team could be performed, if at all, under the MFC model. As with the focus groups some individuals viewed these sessions as brainstorming. Consensus recommendations to be made to management were sought but were not always achieved. Throughout the phase participants were told that ultimately management would make the final decisions on what would be proposed to the Unions in bargaining. At the conclusion of the design team meetings, an event was held where management urged employees to inform their coworkers about the positive aspects of MFC. No further meetings were held with employees after this event. The design team recommendations were passed through several levels of management review. By February the Re- spondent had decided on the parameters of the MFC pilot program that it wished to propose to the Unions. Many of the design team recommendations were accepted by man- agement; others were not. II. LEGAL ANALYSIS It is axiomatic that an employer must bargain exclu- sively with the union with respect to terms and condi- tions of employment. This case presents the issue of whether the employer, in formulating its proposals for bargaining, can consult with a very important resource— its own employees. Our colleague has concluded that the National Labor Relations Act forbids such consultation in this case. We disagree. The complaint alleges that the Respondent violated Section 8(a)(1) and (5) of the Act by dealing directly with employees represented by ESC and CNA concern- ing mandatory subjects of bargaining during the design phase meetings. The judge found, and we agree, that the criteria to be applied in determining whether the Respon- dent has engaged in direct dealing under Section 8(a)(5) are enumerated in Southern California Gas Co., 316 NLRB 979 (1995). They are: (1) that the Respondent was communicating directly with union-represented em- ployees; (2) the discussion was for the purpose of estab- lishing or changing wages, hours, and terms and condi- tions of employment or undercutting the Union’s role in bargaining; and (3) such communication was made to the exclusion of the Union. Applying this test, the judge found no violation and dismissed the complaint. We agree with the judge that there is no basis for find- ing a violation in this case. Although the Respondent communicated with its employees, that discussion was not for the purpose of establishing or changing terms and conditions of employment or undercutting any Union efforts to negotiate. The record emphatically demon- strates that throughout the process of developing and refining its MFC model, the Respondent never excluded the Unions. To the contrary the Respondent kept the Unions informed before and during the design phase. And, most importantly, Respondent made it clear that the design phase would ultimately yield only a proposal to be presented to the Unions for bargaining. With respect to this last aspect the Respondent reiterated its commit- ment to bargain. It did so in every communication with the Unions, as well as in its communication with volun- teer employee participants. It clearly stated that, during the design phase, participants would not be engaged in bargaining or setting any working conditions, and that the design phase was not intended to be a substitute for negotiations with the Unions. Consistent with this ap- proach, the Respondent subsequently engaged in bargaining and reached agreement with Local 250—the union representing those employees whose jobs were most directly implicated by the proposed MFC model. In addition, Respondent was willing to bargain with CNA and ESC at the conclusion of the design phase and after the proposal had been reviewed by a number of man- agement review committees. All of the foregoing was consistent with Respondent’s bargaining obligation and its bargaining history. As the PERMANENTE MEDICAL GROUP 1145 judge noted, the Respondent has had a longstanding bar- gaining relationship—50 years—with the Unions. Our colleague acknowledges that collective bargaining would ensue after the design team had made its recom- mendations and Respondent had formulated its proposal to the Unions. Her concern appears to be based on the fact that some of Respondent’s representatives urged employees to support Respondent’s tentative plans, and urged them to persuade others to support those plans. From this concern our colleague posits that, in subse- quent bargaining, the employees would take one position and the Unions would take another. In response, we note that even if some of Respondent’s personnel urged em- ployees to accept the Respondent’s tentative plan, this does not gainsay Respondent’s fundamental purpose (re- peatedly explained) in setting up the design teams, viz, the plan was tentative; employees were being used to perfect it; the result would be a proposal tendered to the Unions. In addition, the scenario envisaged by our col- league never happened. There is no evidence that, in bargaining, employees took a position contrary to their Union and tried to undercut it. Similarly, our colleague asserts that a union’s bargaining proposals “may reflect” employee desires because the employer has shaped those desires. Again, this assertion is based on speculation concerning what may happen. It is not based on record facts.1 Our dissenting colleague also uses a theoretical con- struct that simply has no application to this case. Thus, she treats this matter as if the Respondent were charged with “dealing with” an employee committee that consti- tutes a labor organization for purposes of Section 8(a)(2). She does so by characterizing the design team as a rival entity for the direct exchange of proposals, thereby by- passing the Unions. She portrays the relationship be- tween management and the MFC project teams as an attempt to create an internal schism—a rival organiza- tion—among represented employees and to usurp and frustrate the Unions’ authority to bargain. We find no evidence to support this contention. To begin with, there is not even an allegation that the design team is a “labor organization” within the meaning of Section 2(5), and no allegation that Respondent violated Section 8(a)(2). Further, in light of Respondent’s com- mitment to bargain exclusively with the Unions after a proposal was formulated, we see no effort to have a rival bargaining group. We acknowledge that Respondent referred to the non- participating employees as the “constituents” of the par- 1 Our colleague seemingly recognizes that there is no evidence of undermining of the Union. She says that Respondent’s conduct “neces- sarily” undermines the collective-bargaining representative. ticipating employees. However, as noted, there is no allegation that the design team was a labor organization representing employees. In addition, the participating employees were free to espouse their own views, and the design team was free to make its own recommendations. Our colleague further asserts that Respondent was “dealing with” the design teams. In this regard, she re- fers to the quoted term as it is used in Section 2(5). However, as noted, there is no such allegation in this case. To the extent that she means “direct dealing” as a 8(a)(5) concept, we again note that Respondent was care- ful to preserve the role of the Unions as bargaining repre- sentative. The “direct dealing” cases cited by our colleague are clearly distinguishable.2 In none of them did the em- ployer do what Respondent did here, viz, assure employ- ees that it was simply formulating a proposal to be bar- gained collectively with the Union. Finally, unlike Obie Pacific, 196 NLRB 458, (1972), the instant case does not involve “surreptitious espionage” or “open interroga- tion.” In sum, Respondent simply turned to its employees to assist it in formulating Respondent proposals to the Un- ions. Respondent always made clear that its bargaining obligation ran to the Unions, and it honored that obliga- tion. Based on all of the above, we conclude that Respon- dent has not violated Section 8(a)(5) of the Act. ORDER The National Labor Relations Board adopts the rec- ommended Order of the administrative law judge and the consolidated complaints are dismissed. 2 Summa Health System, 330 NLRB No. 197 (2000); E.I. Dupont de Nemours & Co., 311 NLRB 893 (1993). DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1146 MEMBER FOX, dissenting. I. INTRODUCTION Contrary to my colleagues, I would conclude that Re- spondents engaged in direct dealing with their employees in the design phase of the Member-Focused Care Project (MFC Project) in violation of Section 8(a)(5) and (1) of the Act. I would reach this conclusion because Respon- dents’ conduct of the design phase necessarily tends to frustrate the basic policies underlying the system of col- lective bargaining envisioned by the Act and to subvert the Unions’ ability to effectively represent the interests of the bargaining unit as a whole. II. THE FACTS The facts are fully set forth in the administrative law judge’s decision. Briefly, Respondents,1 with the aid of Anderson Consulting (Anderson), developed the “Gate- ways Project” (Gateways), later renamed the “Member Focused Care Project” (MFC Project), to address declin- ing customer satisfaction with Respondents’ services. In mid-1995 Respondents invited 150 employees, 8 of whom were represented by CNA and 4 of whom were represented by ESC,2 to participate in 7 day-long focus group sessions on Member-Focused Care (MFC). Re- spondents described the focus group sessions and “the Member-Focused Care Project (formerly known as ‘Gateways’)” in the following manner: Please be aware that the enclosed material3 repre- sents recommendations about new ways of working, technology, roles, and activities. Through discus- sions with staff, managers, and unions, the original models will be tailored and modified to fit what is important for the Health Plan members and people of Kaiser Permanente of the North East Bay cus- tomer service area. The purpose of these sessions is multifold. First, we want to ensure that there is a broad-based under- standing of the original Gateways design work and the Patient Care Core Process model. Secondly, we 1 “Respondents” refers to Respondent Permanente Medical Group, Inc. (Permanente) and Respondent Kaiser Foundation Hospitals (Kai- ser), collectively. The Charging Party California Nurses Association (CNA) is the exclusive representative for about 7500 of Respondents’ registered nurses. The Charging Party Engineers and Scientists of Cali- fornia, Local 20, IFPTE, AFL–CIO, CLC (ESC) is the exclusive repre- sentative of some 680 medical technologists employed by Respondent Permanente. There are at least 10 additional unions representing vari- ous units of Respondents’ employees. 2 Many of the participating employees were members of bargaining units represented by other unions, including Service Employees Interna- tional Union Local 250, Hospital and Health Care Workers (Local 250). 3 Project Manager Marlayne Morgan explained that the “enclosed material” was the Gateways design. will be asking for your feedback regarding these de- signs and their implications for equipment, facility, information systems, and general training require- ments. Next, we need to identify any areas where the conceptual design work either has not been com- pleted or needs to be re-thought given market or Kaiser changes in the elapsed year. Respondents further explained their expectations of each employee as follows: Your role is to give the group your best opinion, based on your own views, in addition to your peer’s views. We realize that no one can truly represent all interests, but we encourage you to speak with your constituents and bring as many of their insights to the table as you can. The focus group sessions commenced on July 13. The same day CNA demanded bargaining over Respondents’ decision to implement MFC and its effects, but Respondents refused. Respondents told the Unions and employees that, once the design of MFC was finalized, Respondents would give the Unions a chance to bargain before implementation. In the focus group sessions Respondents proposed cre- ating a new nonlicensed “multi-skilled caregiver” posi- tion, and directed the participating employees to analyze current job functions and tasks, and to recommend those that could be performed by the multiskilled caregiver. Respondents also asked employees to respond to pro- posed changes in job functions in certain departments. As part of this process Respondent engaged the employ- ees in games to familiarize them with the language and parlance of MFC. CNA director, James Ryder, and ESC representative, Marlayne Morgan, attended several ses- sions, but Respondents barred them from participating in any discussions. Following the focus group sessions, Respondents di- rected the participating employees to inform their co- workers about what had been presented “to educate them” about MFC. Respondents admonished the em- ployees “not to put a negative slant on it, to keep open minded when [they] discussed these things, not to just leave [their] binders around for, you know, viewing [by] anybody because it could be interpreted in the wrong manner.” In late August Respondents notified the Unions and certain employees that it was going to begin the “design phase” of the MFC Project, the objective of which was “to design member focused care positions and processes to meet our patients’ needs.” Respondents stated that the design phase was not a substitute for bargaining with the Unions. PERMANENTE MEDICAL GROUP 1147 In the design phase Respondents grouped supervisors and employees into six teams led by Respondents’ man- agers and Anderson personnel. The teams, which met twice a month from October through December, dis- cussed in detail specific tasks performed by employees in various job categories and attempted to streamline or reallocate those tasks under the principles of MFC. Re- spondents encouraged the teams to reach consensus, but did not require unanimity. On each issue the consensus or majority view became the team’s recommendation to management. Nancy Carlson, Respondent Permanente’s assistant leader for regional nursing, led the care partner team. Also, on the care partner team were managers Scott Morgan and Michelle Schumacher, two Anderson con- sultants, and approximately 30 nonsupervisory employ- ees, including ESC member Kristy Sparks. The care partner team discussed licensing and certification re- quirements, job functions, analyzed various tasks to “eliminate the number of players,” and recommended reassigning existing job functions to the new multiskilled caregiver positions of “care partner” and “service part- ner.” The care partner team also discussed whether specific departments, such as the respiratory therapy department, should be centralized or decentralized. Employee Dal- berti recounted the debate over that department as fol- lows: [W]e kept going around in circles with the leader of our group, Nancy [Carlson], and Georgia Lee and Kris Rudd were the other two. They wanted to decentralize it, they wanted to give some of the respiratory duties to the care partner. And we fought on this and went back and forth, and at one point it must have gone on for an hour, it just seemed like a moot point after a while, and I raised my hand an objected. I said, you know, you’ve asked us to come to some sort of consensus and some sort of agreement, and as a group we have done that. And we tell you we want the department to remain cen- tralized, and you aren’t listening, you just keep telling us that you want it decentralized and that’s not what we’re telling you. And Nancy became angered, I mean visibly an- gered, she kind of tossed her eraser down for the board and pounded her pen on the table and said, okay, well, let me get this straight, you guys want it to be a centralized department, out of the ICU, and we said yes, and the discussion was dropped. Carlson also tried to convince the employees to accept de- centralization of the respiratory therapy department by don- ning a hat with a string through it and admonishing employ- ees that they “needed to Mental Floss [their] cobwebs be- cause [they] were still thinking the old way, and [they] needed to think the new way of the Member Focus group.” The design phase ended in late December, at which time Respondents said they would make final decisions on the teams’ recommendations and then submit the final MFC design to the Unions prior to implementation. Meanwhile, in mid-January 1996 at a wild-west theme celebration for the design teams, Respondents “urged [the employees] to present the final designs and recom- mendations on MFC to other employees.” Carlson told them “to scout the new territory and tell other employees about the positive aspects of MFC.” Respondents then completed their review of the teams’ recommendations and adopted many, but not all, of the recommendations. Respondents then met with CNA and, on one occasion with ESC, to discuss the finalized MFC proposal. No agreement was reached with either union. III. THE JUDGE’S DECISION The administrative law judge dismissed the complaint in its entirety. The judge found that Respondents’ pur- pose in communicating with employees was “to formu- late a knowledgeable proposal and present it to the Un- ions,” analogizing the design team sessions to question- naires seeking suggestions from employees on how to improve efficiency. The judge also found that Respon- dents did not intend to undercut the Unions’ exclusive bargaining authority, relying on Respondents’ statements that MFC would be bargained with the Unions prior to implementation. Finally, the judge concluded that there was insufficient evidence that Respondents attempted to determine or influence employee sentiment regarding MFC. I disagree with each of these findings. IV. ANALYSIS An employer violates Section 8(a)(5) and (1) of the Act by refusing to bargain collectively with the exclusive representative of his employees. The duty to bargain as defined in Section 8(d) of the Act mandates that the par- ties act in good faith. For the employer, this duty “re- quires at a minimum recognition that the statutory repre- sentative is the one with whom it must deal in conducting bargaining negotiations, and that it can no longer bargain directly or indirectly with the employees.” General Electric Co., 150 NLRB 192, 194 (1964), enfd. 418 F.2d 736 (2d Cir. 1969), cert. denied 397 U.S. 965 (1970). In short, “the employer’s statutory obligation is to deal with the employees through the union, and not with the union through the employees.” General Electric, 150 NLRB at 195. Accord Summa Health System, 330 NLRB 3, 5 (2000). Finally, in applying these principles, I empha- size, as the Board has in cases involving employee par- DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1148 ticipation programs, that we must “focus[] on the evi- dence showing what the organization actually does,” not the employer’s stated intentions. Polaroid Corp., 329 NLRB 424 (1999). The MFC Project at issue is a program to revamp the delivery of health care services in Respondents’ hospi- tals. Notwithstanding Respondents’ contention that de- sign phase did not involve, “in any significant way,” mandatory subjects of bargaining affecting CNA or ESC represented employees, even the judge acknowledged that the teams “may have” discussed conditions of em- ployment. Indeed, at Respondents’ direction, the design teams discussed new job classifications, which aspects of represented employees’ jobs should be reassigned to those new positions, and even the elimination of unit positions. These are clearly mandatory subjects of bar- gaining. See Summa Health System, 330 NLRB 4 (2000); A.M.F. Bowling Co., 303 NLRB 167, 171 (1991), enfd. in pertinent part 977 F.2d 141 (4th Cir. 1992).4 Admittedly, Respondents indicated a willingness to discuss these subjects with the Unions prior to full im- plementation. But, without the Unions’ consent to waive their representational rights, Respondents also estab- lished a mechanism by which they first dealt directly with employees, acting in a representational capacity, about these same subjects of bargaining, and then ad- monished those same employees to present manage- ment’s vision of MFC to their fellow employees in a positive light prior to any bargaining with the Unions. Indeed, the admitted objectives of the focus group ses- sions and design team meetings were to have union- represented employees consider and respond to Respon- dents’ “recommendations” about new ways of working. Moreover, Respondents explicitly encouraged participat- ing employees to survey and represent the views of their “constituents.” At the same time, as this period of rival representation was unfolding, Respondents refused to bargain with the Unions over the same subjects, delaying any discussion with the Unions until after working out the basic terms of MFC with their self-appointed repre- sentative of the employees. In my view Respondents’ conduct necessarily under- mines the collective-bargaining process and the principle of exclusive representation on which it depends. Con- gress embraced in Section 9(a) of the Act the principle of exclusive representation to foster stability in collective- 4 In this regard, I note employee Wendy Dalberti’s unrefuted testi- mony that, in the care partner team’s ongoing discussion of the reas- signment of job functions from bargaining unit personnel to the new member-focused care positions, team leader Carlson eventually ac- knowledged that it was a matter of “who’s cheapest.” bargaining relationships. This exclusivity provision pre- vents employers from positioning one faction of employ- ees against another, or against the majority representative itself. See Medo Photo Supply Corp. v. NLRB, 321 U.S. 678, 683–685 (1944) (“Bargaining carried on by the em- ployer directly with the employees, whether a minority or majority, would be subversive of the mode of collec- tive bargaining the statute has ordained”). In addition, by requiring the employer and the employees themselves to look exclusively to the union to represent the employ- ees’ interests, the Act maximizes the potential effective- ness of that representation by allowing the union to speak with one voice. See Emporium Capwell Co. v. Western Addition Community Organization, 420 U.S. 50, 70 (1975) (union has “legitimate interest in presenting a united front . . . and in not seeing its strength dissipated and its stature denigrated by subgroups within the unit separately pursuing what they see as separate interests”). Here, Respondents frustrated these statutory objectives by directly engaging selected union-represented employ- ees in an ongoing discussion of mandatory subjects of bargaining implicated by MFC, and by directing those employees to meld their opinions and preferences regard- ing those subjects, as well as those of their “constitu- ents,” into recommendations to management. As indi- cated above, the harm in permitting such tactics is that when the union, as the exclusive representative, comes to the negotiating table to represent the bargaining unit as a whole, it must deal not only with the employer, but with the positions staked out by what amounts to a separate labor organization. As a result, the union is unable to present a united front; it is effectively forced to accede to the tentative agreements reached between the employer and the rival representative or to bargain against seg- ments of its own constituency. This situation reasonably tends to produce the ill consequences—factionalism, unrest, and the weakening of the employees’ representa- tive—that, the Congress that enacted Section 9(a) of the Act was trying to avoid by granting unions exclusive representational rights. Moreover, as the General Counsel points out, Respon- dents’ actions trigger concerns similar to those the Board expressed in Obie Pacific, Inc., 196 NLRB 458 (1972), which held that an employer violated Section 8(a)(5) and (1) of the Act by polling employees’ preferences regard- ing a contract clause prior to bargaining with the em- ployees’ representative over revisions to the clause. In finding that the employer’s direct dealing with employ- ees impermissibly infringed on the union’s status as the exclusive bargaining representative, the Board explained: Part of the task facing a negotiator for either a union or a company is effectively to coalesce an admixture of PERMANENTE MEDICAL GROUP 1149 views of various segments of his constituency, and to determine, in the light of that knowledge, which issues can be compromised and to what degree. A systematic effort by the other party to interfere with this process by either surreptitious espionage or open interrogation constitutes clear undercutting of this vital and necessar- ily confidential function of the negotiator. It is indeed designed to undermine the exclusive agency relation- ship between the agent and its collective principals. [Id. at 459.] The present case implicates many of the same con- cerns because Respondents, by directing employees to craft recommendations to management on mandatory subjects implicated by MFC, deprived the Unions of the opportunity to develop, shape, or perhaps even reject, those recommendations in formulating a response to MFC that would fairly represent all bargaining-unit em- ployees. Contrary to my colleagues’ assertion, the fact that Obie Pacific involved arguably more egregious methods of employer interference is beside the point. The violation of the Act lies in the interference itself, whether clumsily carried out by bald-faced interrogation or, as here, deftly achieved through more subtle tactics. In either case, the employer’s conduct necessarily tends to undercut the bargaining representative’ exclusive rela- tionship with the employees and to hinder the union’s ability to effectively represent the bargaining unit as a whole. For all of these reasons, I would find that Respondents failed in the design phase to satisfy a basic element of their duty to bargain in good faith—to deal with the em- ployees through the Unions and not to deal directly with the employees and then with the Unions, as Respondents did. See General Electric Co., 150 NLRB 192, 195 (1964), enfd. 418 F.2d 736 (2d Cir. 1969) cert. denied 397 U.S. 965 (1970). Contrary to my colleagues’ suggestion, my conclusion here would not prohibit an employer from consulting with its own employees in preparing its collective- bargaining proposals. My approach simply recognizes, and adheres to, the line drawn in Southern California Gas Co., 316 NLRB 979 (1995), on which my colleagues and the Respondents rely, between permissible informa- tion gathering and impermissible dealing. In Southern California Gas, the Board found that an employer did not violate the Act by involving its employees in a multi- stage “re-invention” program. The decision emphasized that, unlike in the present case, the employer limited em- ployees’ involvement to collecting data about employ- ees’ tasks and duties, activity that amounted to a “desk audit,” 316 NLRB at 982, and that “none of [the employ- ees] performed the next stages, task valuation and/or idea generation or participated in any recommendation proc- ess.” Id. Thus, the Act does allow employers to consult their employees within certain limits. In this instance, the Respondents simply exceeded those limits. In this regard, the Respondents’ attempt to analogize the activity of the design teams to the information gather- ing activity of the quarterly safety conference approved of by the Board in E. I. du Pont & Co., 311 NLRB 893 (1993) (Du Pont), does not work. In Du Pont, the Board considered whether the employer violated Section 8(a)(2) of the Act by dominating several ongoing employee safety and fitness committees, and whether the employer violated Section 8(a)(5) by dealing directly with employ- ees on those committees and during quarterly safety con- ferences. In finding unlawful domination of the safety and fitness committees, the Board addressed at length the concept of “dealing with”: [T]he term “dealing with” in Section 2(5) is broader than the term “collective bargaining.” The term “bar- gaining” connotes a process by which two parties must seek to compromise their differences and arrive at an agreement. By contrast, the concept of “dealing” does not require that the two sides seek to compromise their differences. It involves only a bilateral mechanism be- tween two parties. That “bilateral mechanism” ordinar- ily entails a pattern or practice in which a group of em- ployees, over time, makes proposals to management, management responds to those proposals by acceptance or rejection by word or deed, and compromise is not required. In an attempt to further define dealing, the Board specifi- cally distinguished “brainstorming” as follows: [A] “brainstorming” group is not ordinarily engaged in dealing. The purpose of such a group is simply to de- velop a whole host of ideas. Management may glean some ideas from this process, and indeed may adopt some of them. If the group makes no proposals, the “brainstorming” session is not dealing. [Id.] We found that the fitness and safety committees had en- gaged in dealing because the employee-participants did not merely list ideas, but made proposals on mandatory subjects of bargaining, and management effectively accepted or re- jected the proposals. 311 NLRB at 894–895. Applying the same concept of “dealing,” the Board also found that the employer unlawfully bypassed the union by dealing directly with the employees on the committees in violation of Sec- tion 8(a)(5) of the Act. 311 NLRB at 897. In short, the Board reasoned that the employer had set up parallel deal- ings with employees on subjects that were identical to sub- jects that were being bargained with the union, thereby un- DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1150 dermining the union’s role as the employees’ exclusive rep- resentative. On the other hand, the Board found that the employer had not engaged in unlawful direct dealing with employ- ees in the quarterly safety conferences. As the Board found, those conferences had proceeded as follows: After supervisors and managers made opening remarks, the conferees broke into small groups to discuss spe- cific topics such as communication of safety informa- tion. They were told that bargainable matters could not be dealt with, that the conference was not “a union is- sue.” In the small groups, employees shared their ex- periences on the topic, stated what they thought the ideal situation would be, discussed what barriers there were to reaching the ideal, how to overcome the barri- ers, and how to implement improvements. [Id. at 896.] In addition, the employees had been told specifically to re- late “their personal experiences,” and “not to act as repre- sentatives” of their work areas. 311 NLRB at 919. On these facts, the Board found no element of “deal- ing” in the safety conferences because they “amounted to brainstorming sessions where the employees were en- couraged to talk about their experiences with certain safety issues and to develop ideas and suggestions.” 311 NLRB at 896. The Board reasoned that the “[employer] did not charge the conference with the task of deciding on proposals for improved safety conditions.” Id. In- deed, the employer had not presented employees with any proposals; it simply asked them to “make whatever safety suggestions they had.” 311 NLRB at 919. The Board concluded that “this style of brainstorming does not constitute ‘dealing with,’” and, thus, could not be considered unlawful direct dealing. Here, contrary to my colleagues’ and Respondents’ contention, the design team meetings differ in important ways from the quarterly safety conferences in Du Pont. Unlike the employer in Du Pont, Respondents did not encourage employees to “make whatever suggestions they had” on how to improve the delivery of health care services at Respondents’ facilities. Instead, Respondents proposed its own model, MFC, for revamping their hos- pitals and employees’ job duties, structured the ensuing debate around that model, and directed the teams to first formulate and evaluate responses to that model and then select one of those responses through consensus or ma- jority decisionmaking and make it their “recommenda- tion” to management. Thus, the teams did not merely produce a list of options, but proposed a course of action. Moreover, Respondents did not instruct employees to relate “their personal experiences” and “not to act as rep- resentatives” of their work areas. 311 NLRB at 919. To the contrary, Respondents explicitly directed employees “to speak with [their] constituents and bring as many of their insights to the table as you can.” This is in stark contrast to what took place in the quarterly safety confer- ences in Du Pont. For similar reasons, several cases cited by the judge, including Logemann Bros. Co., 298 NLRB 1018 (1990), East Tennessee Baptist Hospital, 304 NLRB 872 (1991), and Emhart Industries, 297 NLRB 215 (1987), do little to help Respondents’ case. In Loge- mann Bros., the Board found no violation where an employer only asked employees to complete a general survey containing an “open-ended solicitation for em- ployees’ suggestions on how to improve plant effi- ciency.” 298 NLRB at 1019–1020. Similarly, in East Tennessee Baptist Hospital, the employer limited its contact with employees to collecting data on staffing. 304 NLRB at 872–873. Finally, in Emhart Industries, the employer lawfully gathered employees’ ideas on improving production and quality. 297 NLRB at 225– 226. It may be that some of the design team meetings involved similar efforts to ascertain employees’ daily tasks. But the facts show that Respondents went well beyond this type of lawful data collection. Respondents’ attempt to characterize the design phase as a series of information-gathering sessions is further under- mined by the overwhelming evidence that they actively sought to influence employee sentiment concerning the MFC model. Thus, after the focus group sessions, Respon- dents sent employees off with instructions “to educate” their coworkers about MFC. Moving into the design phase, Re- spondents continually counseled employees to put MFC in a “positive light” when speaking with their coworkers. On the care partner team, Carlson did not simply ask employ- ees if the respiratory therapy department could be decentral- ized; rather, she determined whether the employees wanted the department to be decentralized and then employed vari- ous tactics to convince employees to accept Respondents’ position (that is, make it their “recommendation” to man- agement). Then, in the form of a homework assignment, Respondents asked employees, “what messages do you think should be conveyed to your peers as Member-Focused Care moves forward?” Such a question is plainly calculated to ferret out participating employees’ fears and apprehen- sions about MFC, as well as those of their coworkers.5 Fi- nally, at the old west celebration, Respondents again “urged [the participants] to present the final designs and recom- mendations on MFC to other employees” and “to scout the new territory and tell other employees about the positive aspects of MFC.” 5 In my view, the unlawfulness of this question is not diminished by the fact that Respondents also asked the employees to comment on what they thought managers needed or wanted to hear about MFC. PERMANENTE MEDICAL GROUP 1151 All of the foregoing evidence demonstrates that Respon- dents were not merely gathering information in the design phase of the MFC Project. As the employer in Du Pont did with the ongoing fitness and safety committees, Respon- dents effectively, though perhaps more subtly, used the design teams as an alternative employee representative to pitch MFC to employees, monitor and influence those employees’ feelings about MFC, and, ultimately, to work out the basic structure of an MFC program with those employees prior to any bargaining with the Unions. Respondents’ remaining arguments do little to alter this conclusion. Respondents contend that there could not have been “dealing” in the design phase because the teams were not required to reach agreement on MFC. On the contrary, as the Board explained in Du Pont, “‘dealing’ does not require that the two sides seek to compromise their differences.” There need only be evi- dence of a bilateral exchange over proposals. Based on what happened in the care partner team meetings, par- ticularly Carlson’s attempt to convince employees to accept decentralization of the respiratory therapy de- partment, I would find that such exchanges occurred here.6 Respondents also assert that Du Pont and the princi- ples of “dealing” that have developed under Section 8(a)(2) of the Act do not apply to allegations of unlawful direct “dealing” under Section 8(a)(5) of the Act. But, in Du Pont itself, the Board’s analysis of whether the quar- terly safety conferences involved “direct dealing” in vio- lation of Section 8(a)(5) of the Act hinged on the defini- tion of “dealing” it applied in finding that certain com- mittees were statutory labor organizations for purposes of Section 8(a)(2) of the Act. See also Summa Health System, 330 NLRB 1379, 1381 (2000) (finding that cer- tain management-employee teams “dealt with” the em- ployer for purposes of Section 8(a)(2) of the Act, and, based on the same principles of “dealing,” that the em- ployer violated Section 8(a)(5) by “dealing directly” with employees on the teams).7 Respondents also emphasize their repeated assurances that, after the series of design team meetings and a man- 6 This is just one example. As the judge found, the care partner team covered a host of additional topics, including the “specific job functions of EKG technician, laboratory technician, respiratory therapist, and physical therapy assistant and recommended allocation of these func- tions between service partners and care partners.” 7 I recognize that Southern California Gas Co., 316 NLRB 979, 982 (1995), without citing Du Pont, suggests another test for direct dealing. Without more, though, I am not persuaded that Southern California Gas actually alters the concept of “dealing” laid out in Du Pont, or establishes a meaningful distinction between “dealing with” under Sec. 8(a)(2) of the Act and direct “dealing” under Sec. 8(a)(5) of the Act. In any case, as shown above, Southern California Gas is factually distin- guishable. agement review, they would, and in fact did, submit the final design of MFC to the Unions for bargaining. The judge interpreted these assurances as an indication of Respondents’ good faith. It is, however, no defense to a charge of unlawful direct dealing that an employer is willing to give the employees’ exclusive representative some belated opportunity to discuss a proposal that has already been vetted through the employees. As Obie Pacific indicates, the fact that bargaining is contemplated only confirms that the employer’s prior dealing with em- ployees is intended to undercut, or will have the effect of undercutting, the union’s confidential relationship with and ability to effectively represent those employees. See also Alexander Linn Hospital Assn., 288 NLRB 103, 106 (1988) (employer unlawfully surveyed employees’ pref- erences among health benefit plans prior to bargaining with union), enfd. sub nom. NLRB v. Walkill Valley Gen- eral Hospital, 866 F.2d 632 (3d Cir. 1989). In a similar vein, my colleagues erroneously rely on the absence of evidence that the employees actually took a posi- tion contrary to the Unions. In fact, we do not know if there was such a conflict because the Unions deferred bargaining over MFC pending the outcome of this case. But, in any case, the harm done cannot be determined by whether the Respondents actually caused a rift between the Union and the employees. Indeed, a union’s bargaining proposals may reflect the positions taken by the employees in their prior dealings with the employer precisely because the employer has already shaped the employees’ desires, thereby render- ing futile any attempt by the Union to take a different stance. Rather, the harm done here lies in the Respondents’ interference with the Unions’ exclusive, confidential rela- tionship with the employees, particularly as that relationship concerns the formulation and valuation of bargaining posi- tions on behalf of the unit as a whole. It is that interference and the Respondents’ failure to honor the Unions’ exclusive representational rights that defines the violation of the Act. See Obie Pacific, Inc., 196 NLRB 458 (1972). Finally, Respondents suggest that the small percentage of CNA and ESC represented participants in the design phase prevents a finding that Respondents were attempting to set up an alternative representative to the Unions. In fact, the evidence shows that Respondents used those employees to reach a much broader audience. As described above, Re- spondents explicitly encouraged the employee-participants “to speak with [their] constituents and bring as many of their insights to the table as you can.” Respondents as- signed the employees homework questions that indirectly sought information about other employees’ concerns regard- ing MFC. Further, Respondents urged the employees to “scout the new territory and tell other employees about the positive aspects of MFC.” Thus, although relatively few DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1152 CNA and ESC represented employees participated directly in the MFC Project, Respondents effectively used those participants as a means of measuring other employees’ re- ceptiveness to MFC and cultivating support for MFC among those employees. For all of these reasons, I would find that Respondents violated Section 8(a)(5) and (1) of the Act by dealing directly with represented employees over MFC. Lucile L. Rosen, Esq., for the General Counsel. Thomas E. Geidt, Esq. (Schachter, Kristoff, Orenstein, & Ber- kowitz), of San Francisco, California, for the Respondent. Michael Alden, Esq., of Oakland, California, for Permanente Regional Legal Department. Jonathan Siegel, Esq. (Eggelston, Siegel, & LeWitter), of San Francisco, California, for the Charging Party Engineers and Scientists of California, Local 20, IFPTE, AFL–CIO, CLC. James Eggelston, Esq. (Eggelston, Siegel, & LeWitter), of San Francisco, California, for Charging Party California Nurses Association. DECISION STATEMENT OF THE CASE MARY MILLER CRACRAFT, Administrative Law Judge. This case was tried in San Francisco, California on June 11 and 12, 1996. The charge in Case 32–CA–15032 was filed by Engi- neers and Scientists of California, Local 20, IFPTE, AFL–CIO, CLC (ESC) on October 13, 1995,1 and complaint issued April 17, 1996. The charge in Case 32–CA–15084 was filed by Cali- fornia Nurses Association (CNA) on November 13 and com- plaint issued April 19, 1996. The cases were consolidated by order of April 23, 1996. At issue is whether Permanente Medi- cal Group (Permanente) and Kaiser Foundation Hospitals (Kai- ser or, jointly, Respondents) dealt directly with the employees represented by ESC and CNA concerning mandatory topics of bargaining in violation of Section 8(a)(1) and (5) of the Act. On the entire record, including my observation of the de- meanor of the witnesses, and after considering the briefs filed by the General Counsel and the Respondents, I make the fol- lowing FINDINGS OF FACT I. JURISDICTION AND LABOR ORGANIZATION STATUS Respondent Permanente is a California corporation com- prised of physicians engaged exclusively in the provision of medical services to members of Kaiser Foundation Health Plan, Inc. (Health Plan). Respondent Kaiser is a California nonprofit corporation engaged in providing hospital facilities and services to Health Plan. Respondents and Health Plan have been en- gaged in the operation of a health care system including acute care hospitals and related in-patient and out-patient facilities. During the 12-month period preceding issuance of the com- plaints, Respondents and Health Plan received gross revenues in excess of $250,000 and purchased and received goods and materials valued in excess of $5000 which originated outside 1 All dates are in 1995 unless otherwise indicated. the State of California. During that same period, Respondents each provided services valued in excess of $250,000 to Health Plan. Respondents admit, and I find, that they and Health Plan are employers engaged in commerce within the meaning of Section 2(2), (6), and (7) of the Act and that ESC and CNA are labor organizations within the meaning of Section 2(5) of the Act. II. ALLEGED UNFAIR LABOR PRACTICES Background Respondents provide health care services at hospitals and medical clinics throughout northern California including facili- ties in Vallejo and Walnut Creek, California. CNA represents approximately 7500 of Respondents’ registered nurses and ESC represents approximately 680 medical technologists employed by Permanente.2 In 1993 Respondents and Anderson Consulting (Anderson)3 began developing the Gateways Project (Gateways). Gateways incorporated the concept of “patient-focused care.” After nego- tiating with the Unions around February 1995 Respondents implemented Gateways at a hospital in Fresno, a smaller facil- ity. Gateways has not been implemented elsewhere. In early 1995 Respondents and Anderson began developing “member-focused care” (MFC) as a pilot project. MFC was intended to build on the Gateways model and incorporate its “patient-focused care” concept in a larger hospital setting. Respondents originally considered implementation of an MFC pilot project at the facilities located in Martinez, Vallejo, and Walnut Creek.4 The guiding principles of MFC, set forth in the project overview, are, “Increase patient and family involvement in care; Multi-disciplinary care management; Movement of 2 The parties agree that the following employees constitute units ap- propriate for purposes of collective bargaining within the meaning of Sec. 9(b) of the Act: All employees described in and covered by article I, “Defini- tions,” [licensed medical technologists, now classified as clinical laboratory scientists by state law, at various facilities] of the Janu- ary 1, 1993, through December 31, 1995, collective-bargaining agreement between Respondent Permanente and ESC; excluding all other employees, guards, and supervisors as defined in the Act. All employees described in and covered by article II, “Cover- age” [nurses who can legally practice as graduate registered nurses who are employed to perform nursing service] of the Janu- ary 1, 1994, through December 31, 1996, collective-bargaining agreement between Respondents and CNA; excluding all other employees, guards, and supervisors as defined in the Act. The parties also agree that each of the Unions has been recognized at all relevant times herein as the exclusive representative of these em- ployees by virtue of Sec. 9(a) of the Act for purposes of collective bargaining with respect to rates of pay, wages, hours of employment, and other terms and conditions of employment. The ESC contract which expired December 31, 1995, was extended to January 15, 1996, and a successor contract had been negotiated at the time of hearing, subject to ratification. The CNA contract was effective January 1, 1994, through December 31, 1996. 3 Respondents admit that Anderson is an agent within the meaning of Sec. 2(13). 4 Proposed implementation of the MFC pilot project was eventually limited to Vallejo and Walnut Creek. The Martinez facility was closed. PERMANENTE MEDICAL GROUP 1153 services closer to the patient; Care units based on teamwork; and Simplified clinical and administrative processes.” Development of MFC occurred in various phases. Com- mencing in July, Respondents held “focus group sessions.” In late August the “design phase” began. There are no allegations that the “focus group sessions” violated the Act. However, the consolidated complaints allege that specific “design phase” activities constituted direct dealing concerning the redesign of unit positions and job classifications and/or the reallocation of unit work and the effects thereof. Specifically, the following actions are targeted: convening meetings of participating unit employees; seeking proposals from participating unit employ- ees; making proposals to participating unit employees; attempt- ing to convince participating unit employees of the merits of Respondents’ proposals; seeking consensus from participating unit employees about the proposals; conveying participating unit employees’ consensus to management and relaying man- agement’s responses/counter proposals; and soliciting partici- pating unit employees to “sell” the design phase to fellow em- ployees. The July Focus Group Meetings By May 1995 Respondents invited CNA and SEIU Local 250, Hospital and Health Care Workers (Local 250)5 to attend meetings regarding the development of MFC. CNA initially proposed monitoring any implemented MFC plan by utilizing 10 critical care indicators for a 2-year period in order to deter- mine whether the level of care was superior. This was rejected by Respondents. By letter of July 13 CNA requested negotia- tions over the decision to implement MFC and demanded that Respondents, “cease and desist immediately from discussion currently in progress regarding this issue.” Respondents replied that no decision had been made to implement MFC. Rather, Respondents asserted that the only decision which had been made was to develop a pilot project. Thereafter, CNA represen- tatives attended the July focus group meetings. ESC was not invited to attend these meetings but by the end of May or June, ESC was also aware of Respondents’ plans to develop an MFC pilot project. Although ESC was refused admission to meetings initially, in July, Dildar Gill, labor man- agement consultant for Respondents, phoned Marlayne Mor- gan, senior union representative for ESC, and told Morgan that she would be allowed to attend a 7-day focus group workshop in July as an observer. A letter of July 17 from Ken Dale, direc- tor of labor relations, to Morgan confirmed the invitation and stated Respondents’ intention to bargain with the Union if is- sues affecting unit employees needed to be bargained. From July 13 to July 21, ESC, CNA, and Local 250 attended meetings at the Sheraton Hotel in Concord to discuss the MFC project. Between 100–150 people participated in the meetings which included unit and nonunit employees, management, and Anderson personnel. There were four employees represented by ESC and eight employees represented by CNA included in the focus group meetings. Some of the employee participants were selected to participate in the meetings by Respondents and 5 Local 250 is not a party to this proceeding. However, it was in- volved in MFC focus groups and design phase. others were selected by the Unions. In addition to Marlayne Morgan, James Ryder, CNA director, was allowed to attend the meetings. However, both attended as observers only and under- stood they would not be allowed to actively participate.6 The first 2 days of the July focus group meetings involved an explanation of MFC, how it functioned and provided an oppor- tunity for better patient care. By the third day, employees broke into eight or nine small groups and discussed how the MFC approach might be structured. Employees commented on the current practices and how MFC would change these current practices. These comments were written on a flip chart in the front of the room by the group leaders, who were either man- agement employees of Respondents or employees of Anderson. Several employees testified that in the beginning, group leaders only wrote down the ideas with which they agreed and were compatible with the MFC model. Later, another flip chart was placed in the front of the room to document the concerns of the employees. After employees commented on current practices, the group leaders showed slides outlining the tasks a nonlicensed em- ployee or “multiskilled caregiver” could perform. Employees analyzed jobs, such as a phlebotomist or registered nurse, by each separate task and suggested which tasks needed to be per- formed by a licensed or certified person and which could be done by a nonlicensed multiskilled caregiver. Employees testi- fied that they felt group leaders were trying to pressure em- ployees to put a wide range of tasks into the nonlicensed cate- gory. At the time of these discussions, there were no “multi- skilled care givers” employed at the Respondents’ facilities in Northern California. At the end of the July meetings, employees were told to give information to other employees in their unit who had not par- ticipated in the July meetings in order to educate them on the MFC concepts. Employees were told if they wanted to con- tinue giving input into MFC they should sign up to participate in the “design team meetings” which commenced in October. Communications After the July Focus Group Sessions On July 21 Ryder wrote to Dale announcing that CNA would no longer support the participation of its members in the MFC project meetings. Ryder further stated that the focus group sessions, “confirmed our worst beliefs [that the real purpose of MFC was to assign RN duties to other personnel].”7 CNA de- manded that Respondents stop using employees represented by CNA in further MFC meetings and that Respondents com- mence bargaining over the decision and effects of MFC. As- serting Respondents’ right to talk to employees about work related issues, on July 28, Dale responded by stating that the MFC meetings were merely to receive feedback from the em- ployees and to clarify any mistakes about employees’ job duties that Respondents had made. Dale stated Respondents under- 6 CNA rejected a sponsorship role in development of MFC. 7 On request, Ryder forwarded literature to CNA members regarding CNA’s objections to MFC including an article, “You Play, You Lose,” an article by CNA President Costello identifying MFC as a method of “de-skilling the work force and undermining an RN’s ability to be patient advocates,” and an editorial condemning MFC as detrimental to patient care. DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1154 stood the duty to bargain and would bargain with CNA when there was a decision to implement MFC. On August 1 Morgan wrote to Gill stating that ESC believed its input would not be considered because the Respondents already had “pre-conceived a plan . . . predicated on downsiz- ing of licensed workers and . . . shift[ing] [ESC] members’ duties to lower paid workers in the name of cost containment.” Morgan also reiterated her position that participation of ESC members in MFC design teams and task forces, “does not con- stitute agreement and that any outcome of these discussion that may in any way affect the wages, hours, working conditions or future employment of our members,” be referred to bargaining. Despite these concerns, Morgan agreed to send a list of em- ployees represented by ESC to participate in the October design teams. On August 8 Gill responded by letter to Morgan stating, “[W]e are seeking our employees’ involvement in order to lis- ten to their ideas and the issues they raise.” Further, Gill stated Respondents would provide opportunity to bargain as, “re- quired by law or contract.” On August 23 Morgan forwarded a list of ESC members selected to participate in the MFC design phase project. On August 29 Gill wrote to both Ryder and Morgan enclos- ing a statement which would be provided to all employees asked to participate in the October design team phase. This statement described the task of the design phase as follows: “We are going to design member focused care positions and processes to meet our patients’ needs.” The statement contin- ued, inter alia: You will not be bargaining or setting any working conditions by your involvement in the design phase. This is not intended to be a substitute for negotiations with your union. We are aware of our obligations to discuss changes in working conditions with your representatives. When we have developed a suitable model through this design phase, we will meet and confer with the affected unions as required by law or the contract. Noting that the task of the design phase was to “design posi- tions and processes,” on September 11, Ben Hudanall, business manager of ESC, wrote to Dale stating that ESC would not participate in the design phase because, “the design of positions is a mandatory subject of bargaining.” In addition, Dale noted that ESC believed the purpose of the focus groups, “was to elicit support for predetermined reassignment of tasks . . . in- cluding a transfer of [ESC unit work] to employees in other bargaining units.” ESC withdrew its authorization of the em- ployees it represents to participate in the upcoming design meetings. Nevertheless, Kristy Sparks, an employee repre- sented by ESC, decided to attend the meetings with the under- standing she was not a representative of ESC. The Design Team Meetings In mid-September employees who were to participate in the October design team meetings were sent a letter of invitation by Maureen O’Brien, the Program Director for Respondents. The letter assured employees that their participation was voluntary. Gill met twice with project leaders to ensure that the meetings did not involve bargaining over wages, hours, and other manda- tory subjects. The October design team meetings were held twice a month, beginning on October 6 and ending around De- cember 21. There were approximately eight meetings, each lasting 8 hours, and 180 participants. Participants included Respondents management, Anderson employees and employees of Respondents. Out of the 180 participants, approximately 51 were bargaining unit employees. According to a list prepared in September, there were 21 members of the CNA unit and one member of the ESC unit (the other 28 employees were mem- bers of two other units). Three of the CNA members stopped attending after the first two meetings. Participants were divided into six teams: the care partner team, service partner team, administrative partner team, RN processes/tools team, reaggregation team, and quality manage- ment team. The teams discussed in detail the tasks currently performed by each employee in various job categories and at- tempted to streamline those tasks and eliminate the number of contacts with patients. The teams also noted which tasks needed to be performed by a licensed employee. In addition to this, employees were asked to discuss feasibility of decentrali- zation of some departments. Throughout the design phase, par- ticipants were repeatedly told that ultimately management would make final decisions on what would be implemented and then the matter would go to bargaining. Wendy Dalberti, a phlebotomist represented by Local 250, was assigned to the care partner team. Dalberti testified that employees were requested by Respondents and Anderson to take information back to their coworkers throughout the focus sessions and design phase and, “to keep open minded when we discussed these things.” Within her design team, members dis- cussed whether specific departments should be centralized or decentralized. The employees in her group recommended that the respiratory therapy department remain centralized while the group leaders wanted to decentralize the department. During this debate a group leader, Nancy Carlson, assistant to the re- gional nursing leader, put on a cap which had dental floss emerging from the ear areas. She told employee participants to “mentally floss” as they were thinking the “old way,” i.e., indi- vidual job roles and duties rather than the “new way,” involving care partners and shared duties. Ultimately, the group recom- mended to management that the department stay centralized, although this was clearly not the preference of the group lead- ers. In addition, the team analyzed the specific job functions of EKG technician, laboratory technician, respiratory therapist, and physical therapy assistant and recommended allocation of these functions between service partners and care partners. By December 21 each of the design teams had developed a set of recommendations for presentation to management. Par- ticipants were told that management would make the final deci- sion and the proposed decision would be bargained with the Unions. January Celebration On January 18, 1996, Respondents held a celebration at the Sheraton in Concord for all design team participants. The theme of the celebration was wild west pioneering. A video depicting pioneers going west to conquer new frontiers was shown. Participants were also given small trinkets including a PERMANENTE MEDICAL GROUP 1155 scout card. The scout cards stated that the role of a scout was to blaze trails for a new course and to build bridges in order to make it easier for those that follow. Kristy Sparks, a member of the ESC bargaining unit, recalled that employees were urged to present the final designs and recommendations on MFC to other employees. Nancy Carlson gave a speech in which she told employees to scout the new territory and tell other employ- ees about the positive aspects of MFC. Employees were given an e-mail address and a list of management personnel which they could pass onto other employees who had questions about MFC. In response to employee requests for guidance in answering questions about MFC from coworkers, a list of “Startling Statements” was included in design team members’ packets at the celebration. A group effort was made to formulate responses to these statements. However, employees were told to respond as they deemed appropriate. No further meetings with employee partici- pants were held after the January 18 celebration. Management Process In January 1996 design team recommendations were reviewed by several management groups, including, the MFC steering team, the patient care leader peer group, and an executive group. These groups reviewed design team recommendations and made recom- mendations of their own. Some of these recommendations con- flicted with the design team recommendations. For example, the design team recommended that specialists not be deployed to various team units while the management steering group rec- ommended the deployment of specialists. The steering group also recommended that staff be certified by Respondents em- ployees while the design team recommended training from an outside service. After the first stage of management critique, recommenda- tions of both the design teams and the management steering group were reviewed by a 5-person executive committee of Respondents management. The executive committee was shown the recommendations made by the design teams and management steering groups. The executive committee made their recommendations. Finally, the executive committee’s recommendations were reviewed by top nursing managers. This review period by management lasted only a short period of time from the middle of January to early February. By early February management had a proposed MFC pilot project it wished to implement after bargaining with the Unions. Many of the design team recommendations were accepted by man- agement; others were not. For example, the design teams did not reach a consensus on adding nutritional services and vari- ous cleaning components to duties of a service partner. Man- agement included all of these components in the duties of a service partner. Bargaining with CNA over MFC Meetings between the Respondents and CNA were held on April 25, May 13, and 21, 1996. Euell Winton, senior labor relations representative for Health Plan, attended these meet- ings while Maureen O’Brien, the regional patient care leader, was the chief spokesperson for Respondents. Winton testified that the parties discussed the staff nurse-two job description, training modules, the modular approach to training under MFC for all members of the care team, and skills assessment in order to determine appropriate training. At the last meeting, Winton presented a proposed agreement regarding notice of implemen- tation of MFC, dispute resolution for issues arising during im- plementation, and skills assessment. No agreement was reached according to Winton although he had responded to information requests from CNA during the course of the meetings. Winton recalled that Ryder stated that he was not prepared to bargain about MFC but would attend the meetings to receive informa- tion pending the results of this case. Bargaining with ESC Winton met with Michael Aiden, a business representative for ESC, on April 15, 1996, to discuss the decision and effects of MFC. Aiden asked questions about point of care testing, RN training in phlebotomy techniques, unlicensed test training, and care partner entry of information into an automated laboratory data system. Winton provided answers to Aiden in a letter dated June 4, 1996.8 No further discussions have taken place. Aiden mentioned the unfair labor practice hearing in this case during the April 15 meeting with Winton. Bargaining with Local 250 Winton began bargaining with Local 250 regarding MFC in February 1996. Nine meetings over a period of 3 months in- volved negotiation on job descriptions for six positions repre- sented by Local 250 affected by MFC. Tentative agreement was reached on May 23 and ratification occurred on June 7. Implementation of MFC At the time of the hearing there were no job-related aspects of MFC implemented at Walnut Creek or Vallejo. Administra- tive documents created under MFC, such as the service line and documentation service, have been used in the two hospitals since March 1996. III. ANALYSIS AND CONCLUSIONS Statutory Background Section 8(a)(5) provides that an employer commits an unfair labor practice by refusing to bargain collectively with the ex- clusive representative of its employees. The duty to bargain is defined in Section 8(d). The obligation to bargain in good faith requires, “at a minimum recognition that the statutory represen- tative is the one with whom [the employer] must deal in con- ducting negotiations, and that it can no longer bargain directly or indirectly with employees.” General Electric Co., 150 NLRB 192, 194 (1964), enfd. 418 F.2d 736 (2d Cir. 1969), cert. denied 397 U.S. 965 (1970) (relying on NLRB v. Insurance Agents, 361 U.S. 477, 484–485 (1960), “the duty of manage- ment to bargain in good faith is essentially a corollary of its duty to recognize the union”). 8 Morgan testified on rebuttal that she did not believe the meeting could have occurred between Aiden and Winton because she would have known about it and Aiden was assigned to another case and had no authority to bargain on MFC as far as she knew. Yet, Morgan had no personal knowledge that the meeting did not occur and no other evidence was presented to contradict Winton’s testimony. Aiden was out of the country and unavailable to testify. I find that the meeting did occur. DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1156 Direct dealing is shown by proof that an employer (1) com- municated with represented employees; and (2) that the discus- sion was for the purpose of establishing or changing wages, hours, and terms and conditions of employment or undercutting the Union’s offer to establish or change them; and (3) such communication was made to the exclusion of the Union. Southern California Gas Co., 316 NLRB 979, 982 (1995) (rely- ing generally on Obie Pacific, Inc., 196 NLRB 458, 459 (1972)). Generally, an employer may communicate with its repre- sented employees for legitimate business reasons when it is clear from the outset that any bargainable issues will be taken to the union for collective bargaining and there is no attempt to structure the communication as a bilateral mechanism for mak- ing specific proposals and responding to them. See, e.g., Loge- mann Bros. Co., 298 NLRB 1018, 1019–1020 (1990) (ques- tionnaire to all employees (not just bargaining unit employees), asking for suggestions to improve operations at the facility, such as better ways to perform jobs and use machinery moti- vated by legitimate business concerns.); East Tennessee Baptist Hospital, 304 NLRB 872 (1991), enfd. in relevant part 6 F.3d 1139 (6th Cir. 1993) (survey to solicit input on how to resolve staffing issues); Emhart Industries, 297 NLRB 215, 223–226 (1989) (suggestions for improving production and quality).9 For example, in Southern California Gas, supra, the em- ployer instituted a program to determine where beneficial cost savings might be utilized. The first step, data collection, was alleged to be unlawful direct dealing. Three of 173 unit em- ployees were involved in data collection. Their only function was to collect data about their own and their fellow employees’ jobs. To the extent the data collection might have led to bar- gaining proposals, the employer made it clear that it would bargain with the Union. Under these circumstances, no viola- tion was found. In E. I. du Pont & Co., 311 NLRB 893, 896–897 (1993), the employer held quarterly all-day safety meetings. Employees were told that no bargainable matters could be dealt with and that the conference was not “a union issue.” Employees shared their experiences on each safety topic, stated what the ideal situation would be, and discussed what barriers existed in the workplace to implementing the ideal solution. Comments and suggestions were then forwarded to management. The Board found that these conference were brainstorming sessions and did not constitute direct dealing. The Board noted in particular that Respondent clearly recognized the union’s role at the be- ginning of each meeting. However, an employer may not attempt to erode the union’s bargaining position by efforts to determine employee senti- ment. Obie Pacific, Inc., supra (1972) (polling employees re- garding their views on bargaining proposal interfered with un- ion’s exclusivity); Harris-Teeter Supermarkets, Inc., 310 NLRB 216 (1993) (during initial bargaining, seeking employee 9 See also United Technologies Corp., 274 NLRB 1069, 1070–1071 (1985), enfd. 789 F.2d 121 (2d Cir. 1986) (survey to determine whether personnel policies and benefits were being properly communicated to recent hires held lawful); Vons Grocery Co., 320 NLRB 53, 56 (1995) (seeking preview of upcoming ratification vote held lawful). sentiment on unilaterally changed work schedule before pre- senting it to the union and without admonition regarding bar- gaining with the union, was unlawful direct dealing).10 Analysis Essentially, in response to declining growth in enrollment in Respondents’ Health Plan and rising costs in providing medical services, Respondents produced a draft pilot project and asked employees for feedback and recommendations based on their job experience. Although the design phase required voluntarily participating unit employees to provide information to Respon- dents regarding redesign of certain job positions,11 and realloca- tion of that work, under the circumstances herein it was permis- sible for Respondents to collect information from represented employees about their job duties and to elicit recommendations and solutions from them for problems which existed in the plan design. I do not find that the purpose of the communication with unit employees was to establish or change job duties. I find the purpose was to formulate a knowledgeable proposal and present it to the Unions. At every step Respondents advised the employees that bargaining with the Unions would take place when a final plan was formulated. Less than one percent of unit employees were involved in the design phase and these employees attended voluntarily. Moreover, I do not find a purpose to undercut the Unions. This is not an initial representation case. The parties have en- joyed a bargaining relationship for over 50 years. Respondents sought no bargaining advantage in convening the design groups. Rather, the evidence indicates only that Respondents sought the best proposal to send to management and eventually to forward to the Unions. Finally, it is not fatal to this analysis that the design team sessions may have involved discussion of terms and conditions of employment. As the Board stated in E. I. du Pont & Co., 311 NLRB at 897, “the good-faith effort to separate out bargainable issues and the assurances that the Union had the exclusive role as to such issues are further indications that there was no un- dermining of the Union’s status as the exclusive representa- tive.” The complaints also allege that the discussion in the design team meetings constituted direct dealing because Respondents required employees to come to a consensus about Respondent’s proposals and this consensus was conveyed to upper manage- ment and then relayed back to the participating employees. I do not find that a bilateral mechanism of communication ex- isted and, accordingly, dismiss this claim. Wendy Dalberti, an employee represented by ESC, testified that on a majority of 10 See also Northwest Pipe & Casing Co., 300 NLRB 726, 733 (1990) (show of hands poll to determine support for union’s bargaining position unlawful); Alexander Linn Hospital Assn., 288 NLRB 103, 106 (1988), enfd. 866 F.2d 632 (3d Cir. 1989) (survey of medical, dental or pensions preferences before negotiations unlawful). 11 It is not entirely clear that the care partner, service partner, and administrative partner designs under discussion were bargaining unit positions in either the ESC or CNA unit. According to Respondents, the ESC unit would not be affected at all while the CNA unit would be slightly affected. Employees, however, feared that the positions in both units would be affected. At this point, the issue is somewhat specula- tive. PERMANENTE MEDICAL GROUP 1157 the issues, employees were unable to come to a consensus. Of 30 to 40 individuals in her group, usually only 25 could come to an agreement on any one issue. Scott Morgan, a project manager for MFC, stated if the group could not come to a con- sensus, the majority’s view was recorded along with the ideas and concerns of the dissenters. Both Kristy Sparks and Wendy Dalberti testified that once they gave their recommendations to management, there was no further communication. While evi- dence was presented that employees could access an e-mail system and there voice any concerns or questions they had about MFC, no evidence was presented that employees did in fact e-mail management.12 Additionally, the complaints allege that Respondents directly dealt with employees by soliciting employees to “sell” the re- sults of the design phase to fellow employees. An employer may not undermine the union by discussing bargaining propos- als with employees prior to discussing them with the union in an effort to influence employees regarding the merits of the employer’s proposal and thus weaken the union’s power at the table. Counsel for the General Counsel argues that by pressur- ing employee participants to sell MFC to fellow employees, Respondent has preempted bargaining. Kristy Sparks testified that she was told, “[W]e are starting on the new frontier of a new way of patient care, and [Respon- dents] would like us to fill in the information to all our other employees on the concept of [MFC].” In response, she made MFC materials available to the other employees in her depart- ment. Wendy Dalberti, a Local 250 employee, also testified, “We were told, when we took [MFC] back to our coworkers, not to put a negative slant to it, to keep open minded when we discussed these things, not to just leave our binders around for . . . viewing to anybody because it could be interpreted in the wrong manner. We were to discuss things with them.” Dalberti made the MFC materials available to her coworkers and de- scribed the meetings to the employees in her department. She testified that she did not try to sell anyone on MFC but rather just told them what she knew.13 This evidence falls short of a 12 Dalberti testified that she spoke with a group of employees in her department after the July focus group meetings and before the October design meetings, in order to determine the interests of the employees in her department. While Dalberti may have informed the employees in her department about MFC, she did not present them with any proposal from management. 13 Dalberti also testified that between the focus group sessions and the design phase, she spoke with employees represented by Local 250 about some of their concerns regarding MFC. When the design phase campaign to influence or determine the sentiments of the 7500 CNA-represented and 680 ESC-represented employees. CONCLUSIONS OF LAW 1. The Permanente Medical Group, Inc. and Kaiser Founda- tion Hospitals are employers engaged in commerce within the meaning of Section 2(2), (6), and (7) of the Act. 2. Engineers and Scientists of California, Local 20, IFPTE, AFL–CIO, CLC and California Nurses Association are labor organizations within the meaning of Section 2(5) of the Act. 3. Respondents did not deal directly with unit employees concerning mandatory subjects of bargaining and scope of em- ployment issues, bypassing unit employees’ collective bargain- ing representatives, during the design phase and subsequent phases of the member-focused care project in violation of Sec- tion 8(a)(1) and (5) of the Act. On these findings of fact and conclusions of law and on the entire record, I issue the following recommended14 ORDER The consolidated complaints are dismissed. began, she voiced these concerns. It is not clear whether she specifi- cally stated the concerns on behalf of a group of employees or merely as her own concerns. Moreover, Dalberti’s bargaining representative, Local 250, has not claimed that any unfair labor practices occurred during the design phase or after. Finally, even were these facts the subject of the instant complaints, this one isolated instance does not rise to the necessary level of proof of a dialogue with employees. 14 If no exceptions are filed as provided by Sec. 102.46 of the Board’s Rules and Regulations, the findings, conclusions, and recom- mended Order shall, as provided in Sec. 102.48 of the Rules, be adopted by the Board and all objections to them shall be deemed waived for all purposes.