Role for a Labor–Management Partnership in Nursing Home Person-Centered Care

Purpose: To investigate how a partnership between labor and management works to change the organization and focus of nursing home frontline work, supporting a transition toward person-centered care (PCC) in participating nursing homes. Design and Methods: Using a participatory research approach, we conducted case studies of 2 nursing homes participating in a partnership between a labor union and a provider coalition. The study was designed to reveal whether and how the labor–management partnership supported PCC and to identify challenges to overcome in the future. Results: The partnership provided training and follow-up support to member homes to implement PCC. Management and worker participants used the partnership as a learning collaborative to acquire PCC knowledge and to share implementation experience. Key ele-ments of the implementation in each nursing home were translation of the larger labor–management partnership to each member nursing home, management innovations that developed and supported PCC, and conduct of union actors in each nursing home that supported PCC while maintaining traditional union protections. Frontline workers exhibited strong engagement in PCC practices. Implica-tions:

Under the rubric of culture change or personcentered care (PCC), providers are attempting to redesign nursing home care processes to make them more supportive of resident autonomy and privacy and to implement changes that make nursing homes more like home (Baker, 2007;Misiorski, 2003;Weiner & Ronch, 2003). This transformation grows from the simple but powerful idea that nursing home care need not dehumanize the residents who receive it. Specific practice changes have emerged from rethinking options for life in the nursing home to support residents' preferences about their wake-sleep cycles, food selection and dining times, mode of bathing, and death and dying (see, e.g., Rader & Semradek, 2003).
But transformation of nursing homes has proved more challenging than might be expected. Many traditional nursing homes remain caught in the medical-custodial model identified by Susan Eaton, which relies on top-down, command-style production of care that leaves frontline staff alienated and residents the objects rather than the subjects of life in nursing homes (Eaton, 2000). It is these frontline workers (especially certified nursing assistants [CNAs] but also food service and housekeeping personnel) who interact directly with residents and carry out prescribed care tasks (Brannon, Streit, & Smyer, 1992). Implementation of practices that enact PCC (consistent assignment, job flexibility, eliciting and following resident preferences) directly affects the job design and working conditions of these staff. As important, frontline workers develop tacit knowledge of their residents' needs and preferences critical to PCC. For these reasons, some observers have stressed that frontline workers must be engaged in adapting and implementing PCC for their nursing home workplaces (Yeatts & Cready, 2007).
To engage frontline workers in transforming workplace practices in other sectors, change agents have turned to unions as organizations represent-ing worker interests. Labor-management cooperation and formal partnerships for mutual gains have been established in several industries (Frost, 2001;Kochan & Osterman, 1994). Examples of these partnerships, formed with varying success and longevity to improve efficiency and quality, include General Motors/Saturn (Rubinstein & Kochan, 2001), Southwest Airlines (Gittell, Von Nordenflycht, & Kochan, 2004), and Kaiser-Permanente (Kochan, Eaton, McKersie, & Adler, 2009). These partnerships work to establish highinvolvement work systems (Appelbaum, Bailey, Berg, & Kalleberg, 2000). Worker and management goals align so that workers manage their own performance to meet overall objectives rather than being subject to close supervision for narrow job performance standards. Frontline workers are empowered to make decisions about day-to-day production activities, sometimes through selfmanaged teams, and may be called upon to participate in redesign of their portion of the production process. To support teamwork, autonomous decision making, and worker engagement in rethinking production processes, these workplace systems emphasize training, including cross-training for related jobs. Workers with the training and authority to act more autonomously need less supervision, allowing reduction in mid-level positions and flattened job hierarchies. In some contexts, unions have resisted this type of workplace change due to concerns about the potential for work intensification and job loss as work is streamlined (Gill, 2009). However, this literature suggests that an organization representing workers can bring workers' commitment and knowledge directly to the task of redesigning work, creating mutual gains. Could such a model work to transform nursing home care? With only 9.1% of the nation's private sector nursing home workforce covered by a labor-management contract in 2008 (Hirsch & McPherson, 2009), the nursing home industry is not highly organized. In some areas, unions are in contentious battles to be recognized as the collective bargaining agents for nursing home workers. But in the New York City metropolitan area and elsewhere, unions representing nursing home workers are a well-established part of the nursing home scene, where they bargain with management concerning wages and working conditions and may join with management to lobby for public payment rates and regulations that support improved working conditions, job training, resident care, and industry sustainability. (For background on labor unions, labor-management relations, and health care unions in New York, see Freeman and Medoff [1984], Budd [2008], and Fink and Greenberg [2009]). In 2002, compelled by a belief that the future of the nursing home industry will depend on high quality residential facilities that are more "like home," leaders of the New York Citybased union 1199SEIU and an association of nonprofit New York metropolitan area long-term care providers called the Continuing Care Leadership Coalition formed a partnership to foster PCC (Table 1). The union-management steering committee leading the partnership effort did not prescribe implementation steps to coalition members but rather developed ways to share knowledge and support PCC implementation in 40 member nursing homes. A joint labor-management entity founded to gather and disburse training monies provided resources for this effort (Training and Employment Funds, Table 1).
After a discussion of methods, this article describes the partnership's strategies for disseminating culture change to member nursing homes, their impacts in two case study nursing homes and prospects for the future.

Methods and Data
We used methods drawn from participatory action research (Dodson, Piatelli, & Schmalzbauer, 2007;Heron & Reason, 2001) and case study methods (Yin, 2002) to document whether and how the labor-management partnership supported culture change and to identify future challenges. Research questions were formulated in consultation with the project's Research Advisory Council, composed of representatives of labor and management. These included the following: How does the partnership operate? What are its essential elements? What are the challenges that must be overcome in the next phase of change? We gathered data to address these questions from archival sources, observations at partnership conferences, interviews with key partnership actors, and field research at two participating nursing homes that have achieved some success in implementing PCC. We chose the two study homes based on interviews with a labor and a management representative from each of eight nursing homes identified by the Research Advisory Council as having already made substantial progress toward PCC. The nursing home selection followed Yin's "replication" approach, where a first case is chosen to exhibit the phenomena of interest and a second case is ex-pected to produce results similar to the first, supporting some generalization. Nursing Home 1 (NH1) is a 350-bed former for-profit nursing home that has recently been annexed by a religiously based health care system. Nursing Home 2 (NH2), a 700-bed nursing home founded in the 19th century, is now the nucleus of an elder service organization adapting to its changing urban community. Management at both nursing homes were considering transformation to PCC prior to the formation of the labor-management partnership, so were ready to commit to conscious, deliberate efforts for culture change with the partnership's first citywide conference in November 2003. In both nursing homes, the union represented all frontline workers: housekeeping, food service, and nursing personnel, including CNAs and registered nurses.
We systematically gathered information from management, union leaders, and frontline workers to determine the range and depth of adoption of the labor-management approach to implementing PCC in 2006. Our interview guide left room for respondents to add additional themes. The human subjects protocol was approved by the Brandeis University Institutional Review Board.
Because the partnership's model allowed each nursing home to select, develop, and implement its own initiatives, there was no preset intervention to investigate. We asked respondents to describe their typical workday, supervision, and coworkers, and whether and how PCC figured in their work. We also asked whether "people have voice in decisions" or "a chance to be leaders," and whether "residents or families get some say in how care is done." Within each home, one unit that was most advanced in PCC practices was selected for intensive case study. Within the study unit, which served 30-45 residents, we attempted to interview all staff from the day and evening shifts. Night shift staff were not included due to resource constraints. Respondents included neighborhood coordinators (a new position, discussed below), nurses, CNAs, dietary workers, housekeepers, recreation therapists, dieticians, and social workers; all senior managers and department heads, including the chief executive officer, chief financial officer, director of nursing, and human resources director; and the union organizer (Table 1). In all, we interviewed 37 individuals in NH1 (27 frontline workers, 9 managers, and the organizer) and 34 in NH2 (22 frontline workers, 11 managers, and the organizer). Although we did not systematically collect demographic information from respondents, the majority of managers and professionals were Caucasian and the majority of frontline workers were non-White. Women outnumbered men 6 to 1 among both management and staff.
Data were analyzed and interpreted using content analysis. Two of the investigators separately coded emerging themes in all the data sources, one using ATLAS.ti and the other using a more traditional manual system. Detailed results of these two analyses (including primary data associated Collective Bargaining Agreement: Contract negotiated between employers and 1199SEIU on behalf of its members that specifies wages, benefits, and working conditions, including provisions specifying hours, job descriptions, termination, grievance procedures, and funds for training. Continuing Care Leadership Coalition: An association of nonprofit and public long-term care providers based in the New York metropolitan area. Delegates: Elected representatives of union membership in each nursing home, who remain employees of the nursing home.
Delegates participate in collective bargaining, represent fellow members in union meetings, serve as a conduit for communication between union leadership and members, participate in implementing the collective bargaining agreement, file grievances on behalf of members. Delegates are called shop stewards in some other unions. Grievance: Formal complaint filed by a worker asserting that the collective bargaining agreement has not been followed; grievance procedures are specified by the collective bargaining agreement. Organizer: Union employee assigned as a service representative for union members in several nursing homes; assists with implementation of the collective bargaining agreement in those homes and in negotiation of succeeding agreements. Training and Employment Funds: Entity established in 1969 by 1199SEIU and its members' employers, who under collective bargaining agreement contribute a portion of total payroll to fund its training and other programs; also accesses government and foundation training grants. Union: 1199SEIU, a local of the Service Employees International Union, which represents service workers in many industries across the nation. The nursing home division of 1199SEIU has 40,000 members in the New York metropolitan area, including nursing assistants, housekeeping and dietary workers, and registered nurses. Employees of 1199SEIU, paid by member dues, carry out traditional union functions: outreach to workers in unorganized enterprises, political action on members' behalf, and servicing current members.
with each theme) were shared among all three investigators, who met to identify and refine themes that were common to the two coding approaches and to discuss and resolve differences in themes and coding. The themes reported in the article are the result of these analyses.

Findings
The citywide partnership between the union and the providers supported momentum for organizational change in each nursing home with activities described in the first section. We then describe the intermediate effects of partnership activities, specifically management initiatives for PCC and union support at the nursing-home level. These in turn resulted in changes in resident care. We observed the effects of these changes in frontline workers' reports of their increased support for the autonomy and dignity of residents. We then consider observations about the expansion and sustainability of PCC in the two nursing homes.

Partnership Activities Provide Momentum for Culture Change in Member Nursing Homes
The partnership carries out two types of activities that support evolution of PCC in participating nursing homes. First, it hosts conferences gathering representatives of all 40 partnership nursing homes and involving all levels of staff. Second, the partnership offers intensive, off-unit trainings for interdisciplinary teams from each nursing home. Representatives of management and labor from the member nursing homes steer the partnership, building further broad-based commitment to PCC. The training fund staff and consultants support these activities, which form the supportive context for the real work of change and take place in the partnership nursing homes.
Partnership Conferences.-Several times a year, the partnership convenes partnership-wide, daylong weekday conferences engaging all participating homes. The conferences disseminate information about culture change and provide a forum for the labor-management partnerships developing in individual homes to share strategies and report progress. Frontline workers, professionals, and managers selected by the partnership committee in each home gather at tables by home in a large hall for plenary sessions and disperse to breakout rooms for workshops on topics such as involving families in PCC, measuring results, and managing conflict. This approach builds cross-functional teams to bring PCC learning back to the homes. The conferences are a mix of group exercise, formal instruction, policy discussion, and pep rally. CEOs, union leaders, and state and national nursing home policymakers give speeches of encouragement and support; teams of frontline workers, professionals, and managers showcase successful projects; culture change experts make presentations; and workshops provide skills enhancement. These conferences were expensive, with hotel meeting space, food, materials, and fees and expenses for keynote speakers and workshop trainers running as high as $350,000 for a 1-day conference for 500 participants.
In-Nursing Home Trainings.-The partnership mounts on-site expert-led training programs for the nursing home staff from specific units, including multiple disciplines and all three shifts. Each nursing home selects the units to receive training and chooses the content areas from a menu of alternatives, for example, customer service (person centeredness), palliative care, and gerontology. Workers receive regular pay for participating in training, and homes are partially compensated for hiring backfill substitutes for the workers in training. The sessions are held off the unit but at the nursing home site for the convenience of participants. The training days are staggered over 2-3 months so that the teams can test approaches on their units on the workdays following each session and return to their next training session with experience to discuss. Between 2002 and 2005, approximately 7,000 frontline workers, professionals, and managers from 24 partnership nursing homes participated in these trainings at an average cost of approximately $300 per worker; about 600 staff from NH1 and 800 from NH2 participated in training.
It is up to each participating nursing home to implement what management and staff learn through the conferences, trainings, and other partnership participation. This was accomplished in the two study nursing homes as they established a labormanagement partnership at the nursing home level, and as both management and labor decision makers took action to support implementation of PCC.

Translation of the Labor-Management Partnership Into Each Nursing Home
Case study interviews show that managers, workers, and union delegates and organizers (Table 1) in each study nursing home drew on partnership momentum, vision, and training activities to create new pathways and styles of communication. In effect, each home mirrored the larger labor-management partnership by developing its own version of a partnership at the home level. Although the legacy of labor-management conflict remained, the new partnerships provided ways to handle conflicts and move forward together on PCC initiatives. Both homes had regular joint meetings of top managers, department heads, the supervisors of the pilot PCC units (called neighborhood coordinators), the union organizer, and several union delegates to develop PCC initiatives. For example, in NH1, joint work on a complaint about how CNAs and food service staff were treated expanded into a broader code of respectful conduct covering tone of voice for communication, talking in public and private, greeting other staff with a friendly hello, answering the phone, and responding to call bells regardless of job title. The committee created a video on the code, which is now shown in employee orientation.
Although knowledge of the details of joint union-management activities seemed to fade as our interviews moved down the occupational hierarchy to frontline workers, almost all workers we asked knew there was joint support for PCC and described effects they experienced: They [management] respect us more now. Some of the residents won't take their meds but I know them, I know how to get them to do it. So now the nurses will listen to us. Before (PCC) they wouldn't have. (NH2 CNA) The union organizer who served members at NH1 and several nearby nursing homes related how she worked with NH1 management to create trust that concerns would be addressed jointly: [We now have] a "safe space" . . . which is promoted and protected by my top leadership and management top leadership. So if I have a concern, my concern will be looked at because there's a relationship that people really take seriously and people are willing to fight to protect.
The conference and training activities were crucial to the sustenance of the partnership between labor and management that was developing in each home.
The partnership is like the vehicle to bring us together so that we're all on the same page. (NH1 manager) At the partnership conference . . . they flipped the organizational chart so that residents were on top, then CNAs, and then at the bottom was management . . . as a support system for the people who care the most and know the most about the residents . . . . When they brought this chart in and showed it, there was crazy applause. (NH2 CNA delegate)

Management Implements Changes for PCC
Labor and management learned together about PCC, but it was management that had to initiate the structural changes suggested by the PCC model. The management in each study home created the new position of neighborhood coordinator, held by a nurse in one of the study units and by a social worker in the other, to lead pilot neighborhoods, made up of two nursing units. Management also set up a new dining system that provides food to residents from steam tables located in kitchen areas of each neighborhood. The neighborhood coordinators instituted classic PCC practices on the front line, for example, developing teamwork, flexing work across job definitions, instilling a "just-do-it" ethic, and supporting staff in accommodating resident preferences in schedules and food. To further locate authority in the neighborhoods, management in both nursing homes shifted decisions in certain line departments (recreation therapy, social work) to the neighborhoods while streamlining administrative oversight. Both labor and management respondents expressed the view that these initiatives were part of management's responsibility to initiate, shape, and maintain PCC: It's because it came from the top that it will or even can work. If it came from CNAs, nurses, social workers, etc, it would never work . . . . Because upper management makes it clear that they are committed to culture change, it frees up staff to be creative-to think outside the box. (NH2 social worker) Both homes were still working out how to give more power to neighborhood coordinators and "flatten" management. When asked what she did, one coordinator replied: The director of nursing in NH2 explained how management supported the neighborhood coordinator (in a unit not selected for case study), who is not a nurse: . . . [She's a recreation therapist], so it was required that she take the CNA training. She manages the unit, but there is a nursing supervisor around the clock.
With more authority at the neighborhood level, the reorganization called for department heads in social work, housekeeping, recreation therapy, and nursing to give up line authority and become consultants in their areas of expertise. However, these department heads were torn between ceding control to the neighborhood coordinators and their continuing responsibility for their "silos." A manager in NH1 said that the coordinators did not trust department heads to make decisions concerning their neighborhoods, and the feeling was mutual: Management also has the responsibility to commit the financial resources necessary for creating PCC, to support training (only partially paid for by external funds), renovations, and staffing sufficient for increased resident interaction as well as care tasks. A centerpiece of PCC in both homes was investment in a steam table in the neighborhood so that residents could be served individually, a change from food service using trays assembled in the central kitchen. This was accompanied by costly renovation of space, new management systems, and new worker roles, particularly for food service and housekeeping workers, who became directly involved with residents' dining.
Staffing levels, a management responsibility, did not always support the ideal of PCC. This came across most clearly in contrasts between the day and evening shifts. A CNA working days in NH1 reported that she liked the flexibility of PCC, which allowed her to spend more time interacting with residents and less on assigned tasks, but evening workers complained when she left them unmade beds: "So when the 3pm shift comes in it should be OK, but really it's not like that. They get mad." CNAs from the evening shift observed that they had no additional staff to accommodate additional demands. Said one, "I haven't seen any changes. We work more."

Union Actors Support PCC
Parallel to management, the union leadership in each home needed to act in new ways to support PCC but also had trouble letting go of old ways. Several aspects of PCC tested the limits of traditional union protections. Choices about working "out of role" were left to workers and supervisors on the units. Managers reported instances of new worker and union flexibility with respect to job content. The human resources director at NH1 said that when housekeeping workers were assigned to neighborhoods for the first time, they were not required to help with tasks beyond their own work. He just said, "You're part of the community, and if you want to help you can." We heard from housekeepers, and observed ourselves, that they helped with dining and interacted with residents in other ways. This PCC-related job flexibility could not have been implemented if union representatives had opposed it.
Scheduling by facility-wide seniority was another traditional contract provision that was being relaxed for PCC. To maintain teamwork and permanent assignment of workers to residents, neighborhood coordinators assigned weekly work schedules and vacations by seniority within the neighborhood. Workers often agreed with the schedule, but when they did not, they could and did fall back on the contract clause requiring scheduling by facility-wide seniority. The human resources director at NH1 said, "The older workers know that clause very well." Better labor-management communication seemed to be changing the subjects of grievances under the collective bargaining agreement (Table 1). One manager cited a situation that likely would previously have led to filing of a grievance. A male resident was often "out of control, hitting, cursing, etc. Rather than filing a grievance, the staff put a petition together, [saying] we can't take this." The staff "worked out a rotation to spread [the resident] around," and implemented it through the labor-management committee. Cooperation between labor and management created the expectation that this problem could be solved by the workers concerned, rather than through a formal process. However, some grievances were filed over PCC issues. For example, to address a problem with availability of clean laundry identified by the PCC committee, management asked CNAs to spend time in the laundry and laundry workers to spend time on the unit to understand each other's jobs. A grievance was filed immediately asserting that workers had been expected to work outside of their job descriptions; it was not initiated by the staff involved but by their union delegate.
Rank-and-file workers knew that their union supported PCC, particularly at the citywide level, even when few knew the details. A nurse at NH2 said she and her colleagues "go to lots of trainings" and conferences set up by the union. She also said that the union "gives us teeth," so that frontline worker opinions are not only elicited but also have some force. But not everyone felt supported: The union and management are friends-[smirking] very good friends. I think they shouldn't be such good friends. They're [the union] supposed to be protecting me-not speaking for management. (NH1 evening CNA) The union delegates could set a powerful example by supporting PCC, or not. The delegate on the target unit in NH2, a dining service worker, was cited as a PCC leader. An administrator said, "She's vocal and has stepped out of her role." She enjoyed cooking favorite foods with residents for special Friday lunches. However, there appeared to be more tension among union delegates and organizers (Table 1) than among rank-and-file workers about moving from conflict to cooperation. Although nearly all workers told us they liked PCC, the union organizers seemed to hear more often from unhappy workers. When we asked an NH1 organizer about the sources of the worker satisfaction we observed, he ignored the positive and focused on complaints, including new ones related to PCC. He said workers complained, "I wouldn't have gotten CNA training if I knew I was going to be cleaning tables," and that to these workers, PCC looked like "doing more with less." Angry members told him "the union is selling us out." This cautious, defensive stand contrasted with another organizer, who articulated how the union supported PCC and how it helped frontline workers to buy in: . . . because we got it [PCC experimentation] written as part of our bargaining agreement. I see it as a very vital tool in problem solving and it has given our members a greater awareness of their responsi-bility as workers in an institution. Not just "I am a worker," but "I have an interest in this, I am a part of this, I am owning this, if it fails it means I fail." In summary, the union supported PCC while simultaneously protecting members who felt that changes threatened their contractual protections. The protections created a safety zone where workers and their representatives could try out new jobs, relationships, and roles. Within the context of shared goals for revitalizing resident care, the union was able to surface resistance from the workers most affected by culture change so that, ideally, the pace and details of transformation could adjust to recognize worker concerns.

Impact of PCC Implementation: Worker Engagement, Transformation of Resident Care
Interview and observational data indicate that PCC implementation, kicked off by conferences and trainings and embodied in the partnership at each nursing home, structural management changes, and evolving responses of union representatives, was reaching frontline workers. The great majority of CNAs, nurses, housekeepers, and food service workers said they wholeheartedly embrace PCC, and this enthusiasm seemed to reinforce and solidify change. The most common and important gains cited were the opportunities to better serve residents and thus to find more satisfaction in their jobs. The impact of the PCC efforts on frontline workers and the effects of their engagement in PCC are illustrated in workers' willingness to "pitch in," to find their voice and contribute their ideas, to cooperate with one another, and in their reaction to the expansion of their jobs.
Pitching In.-One piece of the PCC ethic that frontline workers cited repeatedly was the need to discard the "it's-not-my-job" attitude and embrace the "just-do-it" or "pitch-in" attitude. Workers reported that they got back just as much or more than they gave in the form of help from other workers and positive reinforcement from residents. A housekeeper in NH1 reported that his day started with taking residents to breakfast and helping with dining and that "then my job begins." He was not required to do this extra work with residents but rather volunteered to help with transporting, dining, and recreation. He said he was "paid back" when others helped clean up. Nurses changed too: The workers who were having the most difficult time embracing the just-do-it ethic were recreation therapists and social workers. Their concerns included loss of status, seen in lost office space and in being asked to provide direct care; reduced contact with peers as they were integrated into the work of dispersed neighborhoods; and anxiety about reporting to a neighborhood coordinator from another profession who might not understand their discipline. A recreational therapist in NH2 said her colleagues: . . . are scared and confused, because when you go to training and say, "what will I do?" [the trainers] don't know how to explain it . . . . [One therapist colleague] got pulled to be a [neighborhood coordinator] and had to take the CNA course. They worry . . . if a resident doesn't want to do an activity, will I be cleaning or doing CNA work?
Voice.-The PCC initiative created an atmosphere in which frontline workers could contribute ideas and innovate: If you have a problem, you can feel free to go to your charge nurse, your community coordinator, the DON or even the administration, and they'll listen. Other places are not like that. (NH1 CNA) Another level of voice came from participating in the city-level conferences: We had the opportunity to do a presentation on what we call our career track for some of our CNAs to become unit coordinators. [The CNAs] were very excited about doing the presentation . . . . Normally the unit director or the leader would provide this information. I think the modeling provided by [the partnership] has been really helpful. (NH2 manager) Teamwork.-Workers were encouraged to think about their work with residents and about how the unit operated, and to share information about residents. A CNA at NH2 related how the CNAs decided together about resident assignments. In contrast to other units, "Here we sit down with the other CNAs and decide who gets who." She gave the example of an aggressive Spanish-speaking resident, whom she traded with another aide who spoke Spanish. They informed the nurse supervisor, who approved. Another supervisor said: On the weekend a CNA floated in and the housekeeper knew that the resident was very particular in their needs and wants, and so she spoke to the resident first and then the [floating] CNA assigned to her and [then that CNA] switched with another CNA [who knew the resident]. Now this would not happen on any other unit . . . . That's what I call a self-directed response to the needs of the resident. (NH2 manager) A dining service worker in NH2, who previously had little resident contact in his job on the tray assembly line, related that his job now involved working through a resident's needs with colleagues: The nurse or a CNA will talk about a meal for a resident, and I'll confirm with the dietician . . . . The menu says a mechanical [soft, chopped into smaller pieces in a food processor], but the resident says, no. I'll talk to the dietician, and she talks to the resident, and she says, try her on regular.
Expanded Jobs.-The frontline workers were essentially unanimous that PCC-generated tasks outside their traditional job descriptions were more work, but they reported (with some exceptions) that it was more enjoyable, fulfilling, and effective work: When I come in I [go from room to room to] say "Hi-How are you doing?" I help them feel better . . . . "What do you want to eat? If you need food, just call me." Sometimes residents are hard to reach . . . . They say, "I don't want it." Without patience you leave it alone. With patience you go over the options. Patience-and love them. (NH2 housekeeper) I help toilet the residents . . . . I don't mind at all . . . . [There's] not a big hierarchy on our unit. . . . I trust the CNAs implicitly . . . . I need to rely on CNAs' observations of the residents' behaviors, eating, etc., and they are excellent at communicating that to me . . . . They are integral to how I do my job.
Culture change is more work. It's much more work . . . . Now we have to set up the dining room and help puree the food or whatever they need, so that's . . . adding to our job duties. (NH1 evening CNA) Respecting Resident Choices.-Resident preferences are central to the PCC concept, so all aspects of PCC training and implementation focused on resident voice and satisfaction. The catch phrase for this was, "It's all about the resident." Although we did not interview residents or families, it appeared that staff were committed to eliciting their active participation in PCC. Among staff, choice for residents was identified both as the core of PCC and as a mark of a better place to work, even when it was also a source of additional work. Resident choices most discussed by staff were timing of awakening and bathing and what and when to eat. Respondents told us they were putting into practice the slogan "Humans are not a tray ticket" by supporting resident choice about food. A CNA from NH2 contrasted bathing in the PCC unit with bathing in the rest of the home as the absence of "the tug of war." On the PCC unit, staff can "accommodate what they [the residents] like." For example, the evening shift now helps two residents with showers, whereas previously all bathing was carried out by the day shift regardless of when residents wished to bathe. The negotiation of these changes across shifts was facilitated by the participation of all shifts in the unit-based trainings.
Workers who traditionally do not interact with residents and families were supported in developing relationships with these "customers." A housekeeper from NH1 said that before PCC he could "not even talk with residents." He went to the PCC classes and got permission and training to help with food, coffee, and transport. "Now they relate to us-we talk more, some recognition, smiling-look into eyes . . . . I feel that they know me . . . . I benefit-I make someone's day." In another example, a housekeeper from NH2 said that families were sometimes on the phone when he picked it up. "I say, 'I'll get the nurse,' and they say, 'no, I want to talk to you.'" Families and residents who felt their preferences were honored reinforced staff commitment to PCC and helped bolster morale: I learned that teamwork made the job faster. Then you're in a better mood. The residents are more happy too. They're looking to you to make them happy . . . . [I sometimes hear from families], "I'm glad you're here and doing a good job. The place is clean." (NH2 housekeeper)

Prospects for Sustaining, Expanding, Replicating PCC
Despite progress in pilot neighborhoods, no nursing home in this New York metropolitan area labor-management partnership had reached full implementation of PCC by the end of our study. However, the two advanced nursing homes in our study yielded indications of how management, the union, and the two working together can make PCC a lasting reality. These concerned rollout strategy, commitment of resources, identification of difficult problems and their solutions, consolidation of gains, and recognition of unit (and nursing home) individuality.
The advanced nursing homes were able to roll out PCC model components from the pilots into new neighborhoods using off-unit staff training as a facilitator. For example, in the rollout of unit-based dining in NH2, the dining service staff participated in partnership customer service training and then observed on the pilot unit. Sustaining and expanding PCC requires sufficient training resources, including payment for the time that trainees spend off the unit as well as payment for backfill on the unit, and sufficient staffing because it is difficult to implement changes where staffing is tight. An administrator from NH2 described how their rollout transforming food service workers into "dining hosts" fell short because staff had not trained enough around dining host-CNA relationships: The dining staff were eager to get out [on the unit], but when they put the food on the unit, the CNAs backed off. The dining host said, "Where's the help you promised?" Both the union and management were worried that tighter budgets and staff cuts could jeopardize the whole effort.
How do I reach my 2007 objectives with cuts in the rates? . . . . The union could lose religion or at the operational level there could be a disconnect . . . . And management-will management buy in? (NH2 manager) Certain structural aspects of nursing home work can be formidable barriers to transformation, and success requires that they be identified and addressed. One of these is the practice, protected by contract, of scheduling staff by seniority across the nursing home, rather than scheduling at the unit level to maintain consistent assignment of particular staff to residents or groups of residents. Staff with seniority were especially attached to priority in choosing vacation weeks, which they had worked long for. Neither of the study homes had solved this challenge, although NH2 seemed farther along. A neighborhood coordinator at NH2 related the progress they had made but also the remaining agenda: We do a lot more independent scheduling . . . especially for the holidays we work out ourselves . . . .
[But for vacations] downstairs [the Human Resources Department] takes over.
A union organizer at NH1 looked forward to a unit-based future for scheduling: Instead of waiting in line behind 200 or 300 people to speak to one person [the Director of Nursing] about your schedule you now have to wait for 2 or 3 people and you are able to speak to your coordinator who will . . . supposedly . . . have the autonomy to address the issue. So for me that is a plus.
The downfall of many an innovation is "backsliding" or "institutional creep" that pulls staff back to old practices. It is difficult to embed the philosophy of home-like PCC permanently into every nursing home process. An organizer from NH1 suggested the importance of consolidating gains to make sure the PCC approach runs deep as well as wide: I just think that we are at a crossroads at the moment . . . . Instead of keep moving on, I think it would be better if we stopped along the way and perfected some of the things that we have implemented.
Finally, numerous respondents cautioned that there is no cookie-cutter approach to developing and replicating PCC, even within a single nursing home. A training director told us, "You have to grow your own." Because staff and resident personalities and capabilities vary by unit, each neighborhood will take a slightly different path.
As constrained resources are further tightened, the labor and management members of the partnership face the challenge of bringing along their own external constituencies and resources, for example, state funders of training. They plan to continue citywide and nursing home-level training, and to develop ways to support PCC practices directly in the collective bargaining agreement. The 2007 contract includes language to accommodate PCC experimentation in job descriptions and work rules. In May 2007, the partnership launched a 16-nursing home demonstration of new PCC initiatives, supported by the training fund, that use interest-based problem solving to approach labormanagement conflict.

Discussion
Exponents of culture change have enthusiastically described innovative "best practices" and researchers are beginning to evaluate their impact on resident well-being (Rahman & Schnelle, 2008). But it is difficult to transform the organization and content of work in established enterprises, and culture change is far from full adoption in the nation's nursing homes even where this is a goal. The labor-management partnership studied here supplied distinctive supports for such change. First, enterprises committed to change joined to disseminate and reinforce learning about change as it evolved. The partnership model linked the participating nursing homes into a learning collaborative explicitly encompassing worker as well as management learning. The training resources consolidated from the collective bargaining agreement and from state and federal programs would have been challenging for individual nursing homes or even a provider coalition to secure and have supported an effective model for shared learning.
Second, union participation powerfully brought worker views to the table: Traditional union protections meant that changes had to evolve deliberately with input from all concerned, increasing their sustainability. With the confidence that job content, scheduling, and other aspects of working conditions would evolve cooperatively in this "safe space," frontline workers thrived on making contributions to care practices, which were highlighted at partnership conferences and in their own workplaces, and on the recognition and dignity their expanded roles brought them.
As noted previously, the gains of such highinvolvement work systems have been demonstrated for other industries and appear to have a potential payoff in resident well-being here. Work in these nursing homes was being transformed in this direction. Frontline workers' focus on person centeredness aligned their work with that of team members and with management goals. This supported more worker self-direction and a flattening of the management hierarchy. Sparked by group work in training sessions, workers and their supervisors rethought care tasks and reconfigured jobs. Improved communication brought more frontline worker knowledge to resident care.
A qualitative study of a unique partnership has limitations. In New York City, as in some other states and areas, the nursing home workforce has been represented by labor organizations for many years, so the intense conflict often associated with contested labor-management roles is not present. However, in many other areas nursing home workers are not represented by a union; or where they are, traditional adversarial labor-management relationships prevail, precluding partnership for mutual gains. Further, the management of the nonprofit nursing homes in the case study partnership may be more dedicated to improving quality of life for their residents than would for-profit counterparts. Nevertheless, as an innovation to support PCC, the labor-management partnership has the potential to diffuse to other nursing home groups and areas. Although both the partnership and the nursing homes where we studied it are unique, the partnership's evolving experience with a labormanagement partnership for PCC suggests that collaboration at multiple levels (citywide conferences and steering committees, labor-management committees in nursing homes, joint training, and labor and management support for workplace changes) can foster teamwork and PCC practices in work settings. In some states where unions are not active, direct care staff associations have been formed to support frontline worker interests; these organizations could partner with management in a similar way (Paraprofessional Healthcare Institute, 2008).
Our findings point to directions for future innovation and research on PCC implementation by labor-management partnerships. Research should be designed to address the challenges we document, for example, building trust between newly empowered neighborhood coordinators and department heads as the organizational hierarchy is compressed; support of buy-in for PCC from professionals as job content is changed; and balance for staffing and responsibilities across shifts as the focus of work moves from care task completion toward person centeredness. We have documented hard-fought changes in work and care, but future research must determine whether these can be maintained and expanded, and how PCC changes are incorporated into job descriptions and other provisions of the labor-management contract. Most important, although staff respondents told us about how much they valued an increased focus on resident quality of life, future research must measure the impact of PCC innovations on resi-dents' actual experience of autonomy and wellbeing.

Funding
This study was supported by a grant from The Commonwealth Fund to the Training and Employment Funds, a joint program of the League of Voluntary Hospitals and Nursing Homes and local 1199SEIU of the Service Employees International Union.