Abstract
Research recognizes both a tension between standardized work and employee participation as well as the fact that management and labor negotiate both formally and informally over the reorganization of work. Through a comparison of the lean production systems being implemented at three General Motors assembly plants, this article demonstrates the tension between standardization and participation to be socially constructed due, in part, to workplace historical context. Workplace history shapes the attitudes of actors as they negotiate change and can become a significant obstacle to implementing teamwork and employee participation schemes, while perceptions of the future may determine whether or not those obstacles are overcome.
Keywords
Introduction
The reorganization of work has been the focus of a broad literature providing a range of models and workplace practices, as well as important understandings of the complexities involved in efforts to introduce teams and otherwise break down the barriers between the conception and execution of work. Some of this research has been criticized for overlooking the role of labor in shaping outcomes, consequently bestowing on management near limitless power to implement workplace change (Smith, 2001; Vallas, 2006; Hodson, 2001). However, later interventions influenced by Thomas’s (1994) power-process perspective provide valuable insight into workers’ varied responses to, and influence over, the reorganization of work (Vallas, 2003a, 2003b, 2006), as well as the degree to which management might hedge against full implementations to limit their vulnerability to workers’ influence over manufacturing systems (Vidal, 2007; Rothstein, 2006a). These serve as useful reminders that the manner in which work is carried out is ultimately negotiated, both formally and informally, until those doing the work consent to perform their jobs (Burawoy, 1979).
Fundamental to this understanding of the organization of work is that context matters (Brown and Reich, 1989; Thomas, 1989). The organization of work, and its reorganization, represents a settlement between management and labor, each of whom brings to the table their own experiences. Yet, research seldom explicitly explores the history of workplace labor relations on the organization of work. Even less has been said of the manner in which management’s and labor’s perceptions of the future impact the organization of work.
This article demonstrates that workplace history and perceptions of the future interact to influence the organization of work. Through a comparison of the organization of work and labor relations at three General Motors (GM) assembly plants located in the Mexican city of Silao and the US cities of Janesville, Wisconsin, and Arlington, Texas, I demonstrate how a history of contentious labor relations can poison efforts to implement teamwork and participation schemes, but that these obstacles can be overcome if management and labor share a reason for doing so.
The three plants were assembling similar full-size sport utility vehicles (SUVs), and each was under orders to implement GM’s ‘Global Manufacturing System’ (GMS), a lean production process intended to mix standardized work routines with teamwork and employee participation. All three plants had successfully standardized work, but they varied in the degree to which they integrated teamwork and participation schemes, due largely to the respective factories’ histories and prospective futures.
This research focuses on the extent to which each of the plants experienced what has been described as an inherent tension between drives to standardize work and calls for employee participation (Appelbaum and Batt, 1994; Vallas, 2003b, 2006). Where work becomes increasingly standardized, workers become less empowered at their jobs, and their participation becomes less substantive and increasingly consultative (Levine and Tyson, 1990; Vidal, 2007). But merely consultative participation cannot provide the inherent job satisfaction and motivation to participate often attributed to more empowering forms of work (Appelbaum et al., 2000; Kenney and Florida, 1993; MacDuffie, 1995; Womack et al., 1990). In fact, such consultative participation might become associated with increased standardization, thereby creating a disincentive to participate.
However, this research demonstrates that, in practice, this tension between standardized work and participation is not necessarily inherent to the manufacturing systems, but is socially constructed based on the attitudes of management and labor. In Silao, a greenfield site where workers had little industrial experience and a long-term future, the GMS was implemented holistically from the plant’s inception. By contrast, in Janesville, a brownfield site with a history of contentious labor relations and little prospect for longevity, teamwork and participation were only superficially implemented. Finally, in Arlington, where workers and managers shared similar experiences to their counterparts in Janesville, the tension was being overcome as management and labor found common goals to coalesce around and implement teamwork and participation.
This argument proceeds as follows. I begin by reviewing the literature establishing the tension between standardized work and participation, as well as the research enhancing our understanding of the manner in which the organization of work is determined. Taken together, these literatures suggest an ongoing cycle of workers’ expectations shaping the organization of work, which, in turn, affects the job quality that informs workers’ perceptions. I then describe the methods by which this research was conducted before offering a comparison of the GMS at the three plants and the reasons for their differences. The conclusion considers the implications of these findings for our understanding of efforts to implement new forms of work organization.
The Reorganization of Work
An expanding body of literature has sought to understand the nature and implications of new forms of work organization designed to more fully utilize employees’ skills and abilities. Vallas (1999, 2006) places this literature in one of two schools. On the one hand are a group of scholars who see the reorganization of the firm and the workplace as a competitive imperative. The new realities of global competition make obsolete the old bureaucratic hierarchy and strict separation of the conception and execution of work fundamental to the Fordist era. In its wake, post-Fordist organizations gain a competitive advantage through flatter bureaucracies that redistribute responsibilities closer to the point of production, eviscerate the clear line between conception and execution, and harness the full power of workers’ physical labor and intellectual know-how. While the focus of this literature is mostly on the competitive implications of restructuring (Ichniowski et al., 1997; MacDuffie, 1995; Pil and MacDuffie, 1996; Piore and Sabel, 1984; Cappelli and Neumark, 2001; Godard, 2004), the renewed reliance on workers’ skills and intellects is said to create a win-win situation for management and labor in which firms increase productivity and work becomes more fulfilling (Appelbaum et al., 2000; Kenney and Florida, 1993; MacDuffie, 1995; Womack et al., 1990).
By contrast, hegemony theorists contest that ‘win-win’ scenario. Instead, they see in the reorganization of work mostly an effort by management to extract greater labor from workers by manipulating workers in a more sophisticated manner than under mass production (Babson, 1995a; Barker, 1993; Dassbach, 1996; Graham, 1995; Grenier, 1988; Parker, 1993). Labor scholars, in particular, are highly skeptical of efforts to reorganize work that seem to simultaneously weaken unions and intensify the pace of work (Babson, 1995a; Fantasia et al., 1988; Graham, 1995; Milkman, 1991; Moody, 1997; Parker, 1993; Parker and Slaughter, 1988; Parker and Slaughter, 1995; Stewart et al., 2009).
Somewhat lost in these characterizations of the reorganization of work is the sense that workers and managers negotiate over how work will ultimately be performed (Burawoy, 1979). Instead, management is often seen as the organizer of work, often endowed ‘with prodigious organizational capacities’ (Vallas, 2006: 1680), so much so that their failure to do so is noteworthy (Taplin, 2001). Workers and their unions may resist (Hodson, 1996; Roscigno and Hodson, 2004; Vallas, 2003a, 2003b), but the dynamic of negotiation through which a work regime is manifested is rarely explored, and then mostly at the level of formal collective bargaining (Cooke, 1990, 1992, 1994; Eaton, 1995).
More recent work has added important nuance to our understanding of roles and incentives of different actors in shaping workplace change, based on contextual issues. In his comparative ethnographies of steel mills, Vallas (2003a, 2003b, 2006) introduces worker agency by applying a power-process perspective of workplace change as a struggle between actors with often alternative world views. He demonstrates a variety of context-influenced reactions by workers that shape the reorganization of work initiated by management. Using a similar perspective, Vidal (2007) compares six manufacturing plants introducing lean production and finds managerial hedging to preempt worker responses like those Vallas found. Both these studies point to the reorganization of work as an informal negotiation between management and labor, with each heavily influenced by the context.
Perhaps the most documented reorganization of work focuses on the spread of lean production in the auto industry and the variety of forms it takes (Besser, 1996; Mishina, 1998; Adler, 1995, 1999a; Adler et al., 1998; Babson, 1995b, 1998; Carrillo, 1995; Carrillo and Montiel, 1998; Fucini and Fucini, 1992; Kochan et al., 1997; Liker et al., 1999; Rinehart et al., 1997). On the shop floor, implementations of lean production differ widely (Durand et al., 1999; Kochan et al., 1997; Liker et al., 1999), but usually involve some combination and style of teamwork and employee participation schemes, combined with an increasing focus on standardizing work. This combination of standardized work with teamwork and participation raises what Vallas (2006) describes as an ‘ongoing tension’ between the ‘logic of standardization, born of the engineering thrust for predictable, calculable, and closely controlled methods of production’ and ‘the logic of participation … to engage workers’ commitment’ (1690; italics in original). In other words, as work becomes increasingly standardized, the opportunity for ‘substantive’ participation that empowers workers to make decisions declines. Instead, their participation becomes of the limited ‘consultative’ sort. Workers may make suggestions, but are not empowered to make decisions over the way work will be performed (Levine and Tyson, 1990; Vidal, 2007).
The logic of this literature offers the possibility of a virtuous cycle in which substantive participation empowers workers, thereby offering more enjoyable work, which causes workers to further embrace participation. But under what circumstances will workers embrace consultative participation, and how does this affect the organization of work in highly standardized workplaces? The manner in which management and labor negotiate this tension remains little explored.
Methods
While historical context is understood to affect efforts to organize and reorganize work, isolating the role of historical context is difficult. Previous research, even within the auto industry, mostly facilitates broad comparisons of different automakers, producing a variety of products, and implementing a range of lean techniques. Under these circumstances, isolating the effect of historical context on the organization of work is not possible. At best, we get a general understanding that greenfield and non-union sites offer management greater discretion than unionized brownfield sites. However, because all three plants in this study were under orders to implement the same lean production system to assemble either identical or similar Sport Utility Vehicles, plant historical context becomes a powerful variable explaining the differences in their organizations of work.
In the tradition of the extended case method (Burawoy, 1998), this research employed a multi-site combination of ethnographic fieldwork as well as formal and informal interviews. Fieldwork began in Silao and covered roughly six months in 2002 and 2003. Research in Janesville involved several visits a week beginning in the summer of 2003 and stretching into 2005. The Arlington portion of the research was designed to address issues raised by the comparison between the first two plants, and was conducted over a four week period in mid-2006 and January 2007.
At each auto assembly plant, I gained access to the shop floor to observe assembly in process. Details of the way the shop floors operated, as well as local labor relations practices, were culled from conversations with managers, union officials, and workers. Executives in each plant’s human resources department consented to taped interviews, and other members of GM’s management team were interviewed, both formally and informally, often off company property. I also interviewed an engineer involved in implementing lean production in both Janesville and Silao. Further insight into labor relations in Silao were garnered through interviews with nine of the human resource managers in GM’s local supply chain.
Union officials were also accommodating, consenting to formal interviews and facilitating ethnographic fieldwork. The union in Silao proved most open, due in part to their determination to demonstrate that, contrary to the well-earned reputation of many Mexican unions, ‘we are not corrupt’. I frequented their offices and observed them at work. They consented to repeated interviews, some of which I recorded, and provided me with relevant documents. They allowed me to attend official meetings and included me in their social events.
In light of GM’s ongoing downsizing, union officials in the US expressed some reservations about discussing their relationships with plant management, even as they sat for interviews and provided further contacts. Four officials with United Auto Workers (UAW) Local 95 in Janesville consented to taped interviews, and I spent half a day talking to stewards in their office at the plant. Two officials from UAW Local 276 in Arlington agreed to be interviewed, though they requested that I take notes rather than record our conversation.
Likewise, I spoke to dozens of workers who offered candid thoughts on their work, employer, and union. It is difficult to quantify the number, due to the ethnographic nature of this portion of the research. In all three locations, some workers were interviewed formally. Others I engaged in group discussions in local bars. At all three sites there were some who invited me to their homes and became recurring sources of information. Overall, workers in the US were more candid in sharing their opinions of their jobs than their counterparts in Mexico.
So while there was some imbalance in the sources of data at the three sites, I was sure to speak with management, union officials, and workers at each of the plants in order to get a complete picture of the GMS and labor relations at each location. And while there was some significant disharmony between management and labor over the implications of the GMS, there was very little discrepancy in their descriptions of how the three plants operated. Finally, because memories of events can be unreliable, I augmented the recollections I heard of historical events in Janesville and Arlington with the accounts of the day as reported in the two cities’ local newspapers.
The Global Manufacturing System Compared
General Motors learned the GMS from Toyota at NUMMI, their joint venture in Fremont, California. As the Personnel Director in Janesville explained, the GMS ‘is basically the old Toyota production system’, but according to the Plant Planner ‘to put some ownership in it, it’s called GMS – Global Manufacturing System’. As designed, the GMS incorporated all the key components found in the literature of Toyota’s lean production system (Adler, 1995, 1999b; Adler et al., 1998; Besser, 1996; Fujimoto, 1999; Milkman, 1991; Mishina, 1998; Parker and Slaughter, 1988; 1995; Womack et al., 1990): standardization of work, the organization of workers into teams with team leaders, Andon systems to allow workers to stop the assembly line and build ‘in-station’ so that faulty product is not passed down the line, and employee participation programs to solicit ideas for improving the lean system. Of the various components of Toyota’s lean production system, the one being fastidiously implemented at each of the plants in this study was the meticulous standardization of work. Otherwise, the three plants differed sharply in their implementations of the GMS.
Standardized Work
Functionally, the three plants achieved standardization in identical fashion. At each work station in all three plants, the steps of the production process to be performed, and the order and manner in which they had to be completed, were documented as a ‘footprint’. As explained by a manager in Arlington:
Every job that we have in this facility, whether it is in material or whether it is out here on the frame line, there is a standardized listing of every element of that job, and the specific sequence that you are supposed to perform that job in, and the amount of time that is allotted to do that job. And we have that level of detail for every job that is out there. If you are going to move any of those … there is a formalized process that is called an OCP or Operational Change Process, and IE [industrial engineering] maintain those standardized work sheets. Standardized work just basically says ‘here is how you do this job, and you do this job this way every time’. And that is, if you pick up three bolts and put them here – you pick up three bolts and put them there. That is the first thing you do. And you do that the same way every time. Where you run into some quality issues is when people start deviating from their standardized work … The next thing you know you’ve got missed work, or a screw not shot, or something like that because they deviated from the standard work and got out of process. Standardized work is just this is the way you do it and you do it like that every time.
Standardization did not only ensure quality but efficiency. Each task assigned a worker also included a specific number of seconds in which it had to be completed. Careful time studies like those conducted by Taylor (1911), aided by computer modeling, allowed GM to carefully calibrate each line operator’s movement to keep them busy on one car until the next one reached their work station. GM’s goal at all three plants was to keep each line operator busy for 55 seconds of each minute, a not easily achieved feat given the mix of models and options on each assembly line, a factor emphasized by that manager in Arlington:
That’s something that a lot of people lose sight of when you talk about different models that you build. There is option content with all those models …You may do this on a Yukon. Tahoe, that’s pretty similar. Then you throw in a Cadillac Escalade, obviously pretty different. Then you throw in some long wheel base – Suburban and Yukon XLs – they’re different. So you have to balance that. Throw in a four wheel drive, a lot different than a two wheel drive. DVD entertainment system versus no entertainment system, sunroofs, off-road packages … That line moves at a steady rate. Yet every vehicle coming down that line is different … And all that being said, the benchmark is still 55 seconds.
These standardized routines were the key to GM’s lean production system on the shop floor, to which the remaining elements of teamwork and participation were complementary. These elements of the GMS varied between the plants.
Teamwork
Perhaps the greatest change to the shop floor associated with the shift from mass production to the GMS was the reorganization of workers along the assembly line into work teams. Teamwork is meant to collectivize work and responsibility for a portion of the assembly process among a group of workers, typically under the guidance of a team leader who coordinates the team’s activities and serves as the team’s liaison with management. The implementation of the GMS in Silao adhered far more closely to this notion of teamwork than in Janesville or Arlington.
In Silao, teams included six workers, one serving as the team leader. Teams were responsible for a particular ‘zone’ of the assembly line comprising a series of footprints. Throughout the work day, workers rotated among the footprints assigned to their team. Doing so provided ergonomic relief and sensitized workers to the expectations of their coworkers further down the line, as one of GM’s former process control managers who had worked in both Janesville and Silao described:
They rotate their footprints. So, I might start out the beginning of the day at the first footprint within our team, and throughout the day we’re going to rotate those jobs. So, you basically get to do a different job throughout the day instead of the same repetitive thing over and over. That’s neat … ergonomic relief is wonderful. Customer supplier interface is very good – internal customer supplier interface. The part that you’re putting on as it goes down the line, how you put it on, how that work content is done, impacts that later footprint in many cases. So, by rotating, you understand your customer very well, because at times you are the customer and at time you’re the supplier.
In addition to rotating through each job in the team’s zone, the plant’s ‘points of the star’ system required each production worker to also perform an administrative task through which the teams monitored their own productivity, safety record, training of each worker on each footprint, as well as the cost of the parts they used and the scrap they created.
At the center of the star was the team leader responsible for managing and coordinating all the team’s work. Team leader positions were quasi-supervisory. On the one hand, team leaders were expected to work on the line and be prepared to cover the work at any footprint within their zone. On the other hand, they were administrators, coordinating and overseeing the production, training new team members, and communicating with plant management. One line operator summed up the role of his team leader by saying:
He supports everyone. If someone is absent, he steps in. If someone goes on vacation, he steps in … When his team is complete, he’s in charge of making sure everyone has time to do their job, of ensuring quality, of implementing plans to eliminate discrepancies or defects, of keeping his team in good condition … He is the sanity of the team, addressing work issues to keep the team productive and quality conscious, financial issues to check the list of materials with the finances to make sure they are in order. This is what the leader does. He supports everything.
By contrast, teams in Janesville were strictly pro forma. While workers were technically organized into teams of five, including a ‘team coordinator’, these were merely groupings of consecutively placed workers on the assembly line who did not rotate positions and had no training in one another’s jobs. Instead, through a plant-wide seniority system, workers claimed ownership over a specific footprint, and could transfer to a different job up to three times a year. Workers in Janesville intended to gradually move to easier work on the assembly line as they aged, eventually landing a job off the assembly line altogether. As one woman put it, job rotation ‘would ruin the whole seniority system. The idea of the seniority system is that you get to work less and you get easier jobs as you get older’.
Furthermore, unlike team members in Silao, workers in Janesville did not perform any administrative tasks and had no responsibilities besides performing the standardized routines included in their footprint. Within this system, the primary responsibility of team coordinators was to fill in for workers absent for the day or who needed an unscheduled break. And unlike in Silao, where team leaders were selected based on merit, in Janesville they chose the job through the seniority system, often as a way of seguing off the assembly line.
Teamwork in Arlington took a third form. Teams included six line operators and a team coordinator. A few teams had implemented job rotation for ergonomic reasons, but for the most part, as in Janesville, workers were responsible for only the specific job they selected through the seniority system. Yet, teams were far more functional than in Janesville, in part because the local seniority system in Arlington promoted team stability, collegiality and cooperation. Unlike in Janesville, the seniority system in Arlington was team focused. Vacant jobs were first available to other workers within the team on a seniority basis. This team stability and intra-team mobility encouraged informal cooperation. As one team coordinator explained:
Say you and me work together for a year or two. We work next to each other. Maybe we’re working on a tailgate together. I do part of the tailgate and then you come in right behind me and finish it. Well, you’ll know what I do, and I’ll know what you do. So, in case of an emergency, and your team leader is not around, you’re going to cover me or I’m going to cover you.
Furthermore, the role of team coordinators in Arlington was more extensive than in Janesville. They provided unscheduled breaks as in Janesville, but were also responsible for filling in for absent workers. When they did, the team coordinators on either side of the team split responsibility for the team between them, essentially expanding by half the number of workers for whom they were responsible. This required that team coordinators know not only all the jobs in their team but several footprints immediately up and down the assembly as well. In Arlington, team coordinators also ran weekly team meetings, communicated with supervisors on behalf of the team, took the lead in resolving problems on the line, and ordered materials. So while team coordinators selected their jobs through the seniority system as in Janesville, they had less incentive to do so as a way of reducing their workload.
Andon Systems and Building ‘In-Station’
Under lean production, the line operator becomes a guarantor of product quality. The GMS called for workers to insure product quality and build ‘in-station’, never allowing a faulty product to pass further down the assembly line. Building in-station was facilitated through the use of an Andon system that allowed workers to signal for assistance and ultimately stop the assembly line to address problems. In practice, the manner in which Andon systems were implemented differed widely, and generally reflected the nature of teamwork in each plant.
In Silao, the Andon system harnessed the collective knowledge of the entire team, under the guidance of the team leader. Pulling the Andon cord triggered the team’s specific alpha-numeric code on overhead LCD signs throughout the plant and the team’s identifying musical jingle began playing repeatedly. In response, the entire team converged on the worker with the problem. Together, they applied a step-by-step group problem-solving procedure to quickly identify and resolve the problem. Only if workers determined they could not easily address the problem were supervisors contacted.
Just as the Andon system in Silao capitalized on the plant’s fully-functioning teams, the Andon systems in Janesville and Arlington functioned in a manner commensurate with each plant’s style of teamwork. Theoretically, at both locations, a worker experiencing difficulty was supposed to pull the Andon cord to alert their team coordinator, who would then quickly determine whether s/he and the worker could address the problem, or whether a supervisor would need to be called for assistance. In practice, this described the process in Arlington far more than in Janesville.
In Janesville, workers overwhelmingly reported that supervisors routinely violated the protocol by resetting the Andon system and tagging the car for later inspection. In addition, workers repeatedly told of their supervisors begging, exhorting, and demanding they not disrupt production by pulling the Andon cord. ‘Don’t shut me down!’ was the refrain. One woman working in the paint department summed up the prevailing attitude this way:
They can come right through and the foreman will tell you that ‘we’re here for quality. Quality counts. Quality is our number one priority.’ And then we’ll have a breakdown, and all of a sudden ‘don’t stop that line for nothing. If you can’t get it, that’s alright. Just let it go. They’ll get it in the next area. Don’t worry about it. Don’t shut down the line for any reason.’
As a result, workers claimed to mostly ignore the Andon system unless a problem arose that simply prevented them from performing their job at all.
Unlike in Janesville, workers and supervisors in Arlington had reached an understanding regarding the use of the Andon system. Team coordinators took the lead when the Andon cord was pulled and called a supervisor if they needed assistance. If the problem was not fixed by the end of the footprint, and the line stopped, other team members left their posts to try and lend a hand.
Employee Recommendations
Given the emphasis on highly standardized work under the GMS, the role for workers’ input was decidedly consultative, with formal procedures by which workers could submit their suggestions to GM for review. To encourage ideas that saved money, GM offered bonuses of 20 percent of the calculated first year savings that resulted from such recommendations, up to $20,000 for ideas submitted by individuals and $25,000 for those from groups of workers. In addition, local management had discretion in distributing smaller awards. Again, the manner and degree of participation differed sharply among the three facilities.
In Silao, the procedure for handling employee recommendations was designed as an extension of the plant’s system of teamwork. Based on the data teams maintained through their own record keeping and their work experience, line operators submitted recommendations through a program called ‘Ideas and Improvements’. Accepted recommendations were posted on a large ‘kaizen’ display in the center of the plant, including the original submission, a description of the outcome, and before and after photos.
The scope of the recommendations ranged widely. Some were simple ideas for improving job quality, for example by moving a stock of parts to reduce bending. Other ideas resulted in significant savings and bonuses for the workers. For example, at the time of this research, the plant’s internal newsletter was touting one worker who was awarded nearly $8,000 for suggesting the redesign of a shipping container – a tremendous bonus for someone earning $10,000 a year. More systemically, according to one production engineer, workers’ suggestions were fundamental to eliminating the kinks associated with the annual model changes that required engineers to redraw each team’s footprints. As the changes to the footprints were introduced, waves of recommendations for improvements were submitted. These decreased as the teams settled into what appeared to be the most efficient work routines.
In Arlington, though workers’ suggestions were not displayed as predominately as in Silao, the system functioned similarly. Workers submitted recommendation forms. These were forwarded to a team of evaluators who would either accept or reject the idea. And as in Silao, the pace of recommendations peaked with annual model year changes, as the teams began implementing the new footprints.
However, managers and union officials in Arlington emphasized the importance of money more than their counterparts did in Silao. Plant management claimed the Arlington facility was among GM’s top plants in terms of bonuses paid, with roughly $400,000 having been distributed to workers in 2005. Union officials and workers indicated that management was most responsive to ideas that led to cost savings. One human resource manager concurred, indicating that they were most interested in implementing workers’ ideas ‘if it makes business sense’.
In Janesville, however, the program for soliciting workers’ ideas had become ineffective. Workers reported that their ideas and advice went largely ignored, or languished. One line operator reported raising an issue repeatedly over a 10-month period, and that ‘nothing has happened, and it won’t’. A high ranking union official in the plant indicated that when he worked on the assembly line he resubmitted an idea several times over almost an entire year before management acknowledged and implemented his recommendation.
According to the process control manager with experience in both Silao and Janesville, the plant in Janesville was caught in a vicious cycle. Management was skeptical of workers’ ideas and gave them low priority. Workers were doubtful of management’s sincerity in seeking ideas, and would ‘just clog it [the suggestion box] with all sorts of stuff’ that only served to harden management’s skepticism and further delay any action to address workers’ legitimate ideas. Instead, in Janesville the norm was to depend on engineers: ‘It’s basically, “ergonomics – well we’ve got ergonomics engineers out there.” If you have a problem you get hold of an ergonomics engineer and study the problem independent of the person actually doing the work.’
Context, Labor Relations and the Reorganization of Work
While it is true that, with regard to teamwork and participation, the three plants operated very differently, it is important to keep in mind their base similarity. At all three, the heart of the GMS was the strict standardization of work. None of the plants practiced substantive participation. At issue is the varying degrees to which teamwork and the consultative participation called for in the GMS was actually implemented. Previous research demonstrates the importance of labor and management’s preferences, brought to bear through negotiation.
The three factories herein illustrate the importance of plant history and prospects in shaping those preferences and eventual outcomes. Certainly, the plants offer a stark comparison of the organizational possibilities inherent to greenfield sites versus the inevitable obstacles to change posed by brownfield sites. In Silao, GM implemented its GMS wholesale from the outset, unburdened by any local history and determined to harness its ‘newness’ to build institutions and norms that supported its manufacturing system. In Janesville and Arlington, however, implementing the GMS was a process of workplace change, and the sharp contrast between the Mexican plant and its US counterparts illuminates the influence of plant history on the resulting organization of work. In addition, the more subtle differences between Janesville and Arlington demonstrate the legacy of decisions made years before, as well as the importance that perceptions of the future may play in encouraging management and labor to find common ground.
Greenfield Enthusiasm
The wholesale implementation of the GMS in Silao is best understood as both a product and a part of the labor relations regime GM carefully constructed in the region. In Silao, teamwork and employee participation were part of a broader trend in Mexico whereby automakers shuttered facilities in the traditional industrial heartland around Mexico City to start afresh in new northern plants (Micheli, 1994) where they hand-picked unions willing to engage in a ‘new union discourse’ (Bayón and Bensusán, 1998) that included renouncing labor militancy and embracing new forms of work organization (Carrillo, 1995; Shaiken, 1990, 1994).
GM’s Silao assembly plant replaced one in Mexico City. In Silao, GM selected a regional industrial union to represent their workers before any had been hired. SITIMM (as it was known by its Spanish acronym) espoused a ‘new labor culture’ of management-labor cooperation summed up by one union official who said ‘I love my wife. I love my daughter. I love General Motors. And I love my union. And I see no contradiction in any of that.’
Technically, SITIMM negotiated wages and benefits, though union officials acknowledged having little bargaining leverage over GM. On a day-to-day basis, the union provided services to help workers acclimate to factory work and take advantage of their steady incomes, a function carefully aligned with GM’s own goal of maintaining low labor turnover. However, SITIMM did not interfere with the day-to-day running of the shop floor. The union staffed an office just off the shop floor, but did not have representatives available to workers in the production area. In fact, there was no grievance procedure at the factory.
It was within this context that GM carefully screened and trained their workers. The job application required that perspective employees provide detailed educational qualifications, work history, and information regarding their financial obligations, from how many members of their family they support, to whether or not they rented or owned their home and the amount of the monthly payments. GM’s ideal candidate was a young man with no more or less than a ninth-grade education, no previous factory or union experience, a family to support, and financial commitments that would tie them to the job. Once applicants passed the initial screening, they took a series of written tests to measure basic reading and math skills as well as personality type. They were also observed in mock work environments.
Those who passed this screening then began a 12-week training and assessment period. The first eight weeks were classroom training to teach them the GMS and all its components in Silao, including units on ‘Working in Teams’, ‘Communicating with Others’, ‘Resolving Conflicts’, and ‘Valuing Differences’. Afterward, the recruits joined a team for a four-week probationary period. If they integrated themselves into the team, under the guidance of the team leader, the assessment period ended and the applicant became an employee.
It was within this greenfield context that, in Silao, teamwork and employee participation blended seamlessly with standardized work routines, among a core workforce with no history in manufacturing, and that had been handpicked for the likelihood they would meet GM’s expectations. Teams of workers poised to take advantage of what were by far the best jobs in the area regulated their own productivity under the guidance of team leaders who served a quasi-supervisory role. Employee participation provided GM with ideas to improve production. In short, the GMS operated according to plan in a plant designed around the GMS itself.
Brownfield Skepticism
Just as the wholesale implementation of the GMS was possible in Silao because the plant was new, the versions of the GMS found in Janesville and Arlington reflected the experiences of US auto-workers and their supervisors over their careers – a period of time notable for plant closures, downsizing, and outsourcing. Between 1978 and 2003, a confluence of factors led employment of hourly workers by the US Big Three automakers (GM, Ford and Chrysler) to fall from 667,000 (Barlett and Steele, 1996) to 275,000 (McCracken, 2003). By 2008 the number had fallen to 172,000, with further cuts announced. Import penetration of the US market led to a drop in the market share of US automakers from a high of 90% in the 1970s (Rubenstein, 1992) to 60% (and falling) by 2005 (Rothstein, 2006b).
Within this context of declining market share and shrinking employment, the automakers pressured local unions for changes in work rules on the shop floor necessary to implement lean production. As they eliminated factories, the Detroit Automakers ‘whipsawed’ local unions against one another, forcing union leaders at the plant level to concede work rules or risk losing their assembly lines to plants with more amenable labor relations. What resulted were factory regimes characterized by hegemonic despotism, under which labor ‘makes concessions on the basis of the relative profitability’ (Burawoy, 1985: 150; emphasis in original) of the employer. To save their plants, local unions partnered with management to facilitate the reorganization of work.
In both Janesville and Arlington, roughly half the workforces were so-called ‘GM Gypsies’1 – workers who uprooted to take jobs in surviving GM plants when their own factories closed. Their new workplaces were those where GM had used the threat of plant closure to whipsaw the local union into granting concessions to local work rules so the automaker could reorganize the shop floor.
While this threat of job loss was sufficient to motivate workers to keep the assembly line moving, it was not conducive to teamwork and participation. Many workers expressed deep resentment toward the automaker, like one who complained that, ‘they lie to everyone. They lie to the workers. They lie to their own supervisors. They lie to suppliers. They lie to dealers. And they lie to customers. All they know how to do is lie.’
Though this sentiment was common in both Janesville and Arlington, specific plant histories informed local labor relations and influenced the style of teamwork and participation practiced. In Janesville, the superficial manner in which teams had been constructed and participation practiced reflected both the plant’s previous experimentation with reorganizing work, as well as a related history of threats to close the plant, which became increasingly credible as the factory’s 90th birthday approached. (The Janesville plant closed in 2008 when the demand for SUVs plummeted.) In Arlington, whipsawing had brought a change in the seniority system that stabilized teams, and prospects for long-term production provided sufficient incentive to implement all aspects of the GMS to some degree.
Reinforced Distrust in Janesville
In Janesville, the whipsawing began in 1985, when over 1200 workers transferred to a plant GM was opening in Fort Wayne, Indiana, after the automaker convinced the local union and its members that the Janesville plant might close. The plant manager warned the local Janesville Gazette that ‘several GM plants, the Japanese, and other competitors have improved faster than we have’, and that ‘other parts of the country and world want our jobs’ (Dowd, 1986). The local union began negotiating concessions to work rules and embracing ‘jointness’, GM’s early attempt to introduce teamwork and move away from Taylorist production. By 1986, an agreement was reached that granted management greater flexibility by slashing the number of job classifications from 90 to three and saved GM overhead by converting the 40-hour work week from five days to four (DuPre, 1986).
The following month, GM announced that the plant in Janesville would not be among the 11 plants the company was looking to close (AP and Staff, 1986). Instead, as GM Chairman Roger Smith complemented the Janesville workforce for its ‘super attention to detail and quality and an understanding and cooperative relationship’ (DuPre, 1987b), the automaker began shifting production to Wisconsin from Pontiac, Michigan (DuPre, 1987a), and Leeds, Missouri (DuPre, 1988). In 1989, GM decided to move the Chevy Suburban product line from Flint, Michigan, to Janesville and lauded the local union leadership for ‘working with us and [being] willing to look ahead and take a chance’. Otherwise, ‘we wouldn’t be here today’. For his part, the leader of the local union conceded that he was once the ‘most radical mother’, but that he had to change because ‘there are too many cars and trucks out there and too many plants’ (Nelesen, 1989).
Workers who remembered the jointness program described far more substantive participation than under the GMS, which they attributed to the discretion given their immediate supervisors. One worker hired in 1986 as management replaced those who left for Ft. Wayne explained that, back then:
… if your job was bad, you could complain to your coworkers and band together. And maybe a group of you went to the foreman and told the foreman what you want[ed] to do, and the foreman would say ‘OK, try it.’ … When we hired in, the line foreman had a lot of authority, could make a lot of decisions as far as manpower, job placement, fixing things that he sees as wrong with the process.
But that substantive style of participation was short-lived in Janesville, abandoned in a way that left workers doubting future calls from management for teamwork and cooperation.
Under pressure for greater productivity as the Chevy Suburban assembly line was just beginning production, the jointness program was abandoned. Almost overnight, a new management team was brought in. As one manager who was there at the time recalled:
They took our plant manager and basically walked him out the front door and said ‘stay off the property’. He was the plant manager for five years – Gone. Assistant plant manager and general superintendents – Gone. Superintendents – Gone. Personnel Director – Gone … They brought in a guy [as plant manager] from Linden [New Jersey], and a new production manager from Rockport [New York] who was here for one reason – get the product out the god-damn door. And that was the game. Working together – jointness. They could give a rat’s ass. If jointness got in the way of the bottom line – get out of the way. As far as joint meetings at the plant level, floor, formally – nonexistent. Quality counsels, joint leadership committee – not meeting.
After that, the emphasis on standardized work overwhelmed any attempt at teamwork or employee participation. Under constant pressure to increase productivity or risk a closing of the aging plant, the union and workers in Janesville grudgingly accepted the intensification of work as GM steadily added tasks to each worker’s 55-second routine. But with some workers in the plant having experienced GM’s abandonment of the jointness program, and the so-called GM Gypsies bitter for being forced to move, the teamwork and participation called for by the GMS were incompatible with intensified standardized work routines. Workers did not buy into the new production system and supervisors did not expect or trust them to participate. In fact, the notion that workers should tell GM how to improve productivity, cut jobs, and make work harder struck them as ridiculous. The union Shop Chair, the highest ranking union official in the plant, explained:
What’s involved is giving the people on the floor more input in to the daily running of the plant. They [management] make each area into a little bitty team. That team, their job responsibility is a certain amount of jobs in that area. How can they make them better? How can we resolve quality issues? But management wants to come out and say ‘we need to take a person out. How do you think we can take a person out?’ That’s when we put the flag up. No, that ain’t our job, to tell you how to eliminate people. They got their ulterior motives that are tied in there. They want the people to agree on how they’re going to work harder.
Likewise, managers expected no such participation, and routinely behaved in ways that reinforced workers’ attitudes and assumptions.
Hope in Arlington
If in their implementation of the GMS, the local actors in Arlington had managed to reduce the tension between standardization and teamwork and participation more than their counterparts in Janesville, this was due to both historical factors and a view toward the future. Similar to Janesville, the Arlington plant was a survivor of whipsawing. In 1991, GM announced an impending round of plant closings, and specifically pitted the Arlington facility against the automaker’s other plant assembling full-sized passenger cars in Ypsilanti, Michigan. Though GM denied any attempt to whipsaw the localities against one another (AP, 1991), the company entertained offers of incentives from the local unions and state officials to keep their factories open (Knight-Ridder, 1992). In early 1992, GM announced that the Arlington plant would stay open (AP, 1992).
Among the revisions to the plant’s labor agreement to which the local union in Arlington agreed was a provision that reworked the seniority system within the plant to stabilize teams, which then encouraged teamwork based on familiarity. However, the more thorough implementation of the GMS in Arlington than in Janesville and Arlington is better attributed to the plants’ respective prospects for the future than their histories. Local management, union officials and workers in Janesville knew that their 90-year-old plant’s days were numbered. It motivated them to push productivity, lest they provide GM incentive to shut the factory, but served as a disincentive to attempt real change. The local union president emphasized this, suggesting he and his members would support implementing all the aspects of the GMS found in Silao if GM made a long-term commitment. ‘We can do it’, he insisted ‘if they build us a new plant’.
Attitudes and behavior in Arlington were a little different than in Janesville in part because the plant still had long-term prospects. As at nearly all GM’s plants in the US, the possibility of closure was real, but did not represent the immediate threat posed by the Janesville factory’s age. In fact, since beginning assembly of GM’s then hot-selling SUVs in the mid-1990s, the Arlington plant had become a popular destination for workers whose plants had closed. In 2004, management claimed that half the workforce transferred to Arlington in the previous four years. Those workers received two weeks training before starting work ‘to orient them to what we do here at Arlington’, according to one manager. They then joined a work team among a group of workers willing to help out, many of whom were also hoping Arlington would be the last place to which they relocated.
This prospect of longevity provided incentive for workers, their union, and management alike to navigate the most palatable way to implement the GMS. And so in Arlington, teamwork took on a compromised form. In a similar but perhaps less extensive manner than Vallas (2006) describes, workers shaped the GMS to emphasize support for one another and perhaps make their work a little easier. Team coordinators provided service to team members and responded to their problems when the Andon cord was pulled. As in Janesville, workers avoided telling management how to improve productivity, which they associated with harder work. But unlike in Janesville, they did not eschew the suggestion program altogether. Instead, encouraged by management and their union, workers focused on those ideas that would cut costs, thereby helping to keep their plant competitive while also earning them bonuses.
Conclusion
Organizations of work that empower workers are often claimed to enhance job quality through intellectual engagement, thereby promoting worker buy-in. However, where work is highly standardized, and employee participation merely consultative, no such inherent motivation to participate exists. Under what circumstances, then, will such consultative participation schemes take hold?
This research finds that workplace history and perceptions of the future are key to understanding the implementation of new organizations of work. In Silao, at a greenfield site with no history and a bright future, in a part of Mexico where unions compete for members based on their ability to make themselves attractive to management, the GMS was introduced wholesale among a young workforce with no manufacturing experience who had recently landed the best blue collar job opportunities the region had to offer. Silao not only exemplifies the managerial prerogative available in greenfield sites, the case also provides a baseline from which to measure the degree to which the implementations at the US plants veered from GM’s ideal, and to analyze the role historical context played in shaping the manufacturing systems in Janesville and Arlington.
At those facilities in the US, implementation was complicated by the long histories through which management and labor viewed the GMS. For years, GM had been whipsawing local unions against one another to gain concessions over local work rules. In the process, they created a wave of disgruntled workers who were relocating around the country. Both Janesville and Arlington had been subject to the whipsawing and became home to these so-called ‘GM Gypsies’. Unlike in Silao, at both US plants there was a perceived conflict between the renewed dedication to standardized routines that intensified work and simultaneous calls for teamwork and participation.
In Janesville, GM’s earlier abandonment of the jointness program, which emphasized more teamwork and substantive participation than the newest iteration of lean production, cast a heavy pall over the implementation of the GMS. Line operators and the local union clung steadfastly to their plant-wide seniority system which mitigated against job rotations, and the rhetoric surrounding teamwork and participation was openly scorned. Management, too, hedged, declining to challenge the seniority system or implement any form of GMS training to explain the system. And rather than build the trust necessary for the Andon system to function, workers and supervisors ignored the system unless production became otherwise impossible.
Though the plant in Arlington shared a similar history of whipsawing with the one in Janesville, and included a greater array of transplanted workers from around the country, more headway had been made in implementing the GMS. In part, this was due to the legacy of decisions made years earlier, specifically the renegotiation of seniority rules to promote team stability.
However, the biggest difference between Arlington and Janesville lay in their prospects for the future. Unlike in Janesville, in Arlington the potential for decades of continued production remained if they kept pace with other factories in implementing the GMS, which provided powerful incentive to both labor and local management – some of whom had also moved to Arlington due to plant closures elsewhere. Together, they trained workers and oriented transferees to the local production process before sending them to work. Labor and management encouraged widespread participation by focusing on the financial benefits. None of this was happening in Janesville, where all parties understood the plant to have only a short-term future.
Together, the three plants demonstrate that the tension between standardized work and participation is socially constructed. In Silao, teamwork and strictly consultative participation coexisted and even complemented highly standardized work rather seamlessly in an environment in which GM designed the social relations of production around their vision for the GMS. In Janesville and Arlington, the tension was palpable as labor and management negotiated implementation. In this regard, Janesville was neither a case of absolute management incompetence nor labor recalcitrance. Rather, the implementation of the GMS in Janesville was a settlement, however begrudging, around an implementation that would provide steady production based on the meticulous standardization of work. Similarly, Arlington was not strictly a case of management overcoming a reluctant labor force to implement its vision. Instead, management and labor were building on past experiences and using the future as incentive to negotiate ever fuller implementations of the GMS into place among a workforce with bitter feelings toward its employer. These cases, therefore, illustrate the organization of work as the product of both formal and informal negotiation between management and labor.
The irony is that in Janesville, workers’ commitment to their seniority system and skepticism of teamwork and participation exacerbated the negative impact the standardization of work had on their job quality. While the benefits of teamwork and participation for workers on the lean assembly line appear overstated by those who describe a win-win situation for management and labor alike, within the context of standardized and intensified work, teamwork and participation did offer workers some relief. In fact, assertions among scholars that teamwork and participation serve as the tools of hegemonic control by which workers are implicated in their own exploitation under lean production appear unwarranted. The driving force behind the intensification of work at GM was management’s enhanced engineering capacity, which facilitated an assembly line on which parts tend to fit seamlessly into place, thereby allowing work to be meticulously synchronized whether the organization of work included fully functioning teams as in Silao or operated as a more traditional mass production assembly line as in Janesville. Given these realities of lean production, full implementation including teamwork, job rotation, and the participatory elements that take workers off the assembly line are less stressful for workers than partial implementations that focus exclusively on standardizing work.
However, getting management and labor to implement such practices may require they overcome a vicious cycle of skepticism breeding lack of implementation, which feeds further skepticism. The implementations of the GMS at these three plants demonstrate the limits and opportunities for the reorganization of work posed by the historical context in which it occurs. Certainly, management initiates change and labor responds. However, both do so with regard to the workplace environment and with regard for the broader competitive environment that recognizes the risks and rewards associated with recalcitrance and acceptance. This focus on context also serves as a reminder of the fluidity of the organization of work, and that research only captures a specific moment in time. In Janesville, the union president insisted he could deliver Silao style teamwork if given a similarly new plant and lease on life. In Silao, the factory will age, and workers will gain experience. Their union will need to find new ways to remain relevant to their members once they have adapted to the new work environment. Likewise, the workers and managers in Arlington continued to operate within a context of GM’s further downsizing. The descriptions of the GMS in this research, like all studies of the organization of work, are but snapshots of particular work sites at specific moments in time. Those pictures reflect much broader realities in the lives of the workers and managers that impact the way they ultimately agree work will be performed.
Footnotes
Acknowledgements
I would like to thank Matt Vidal for his thoughtful comments on this article.
Funding
Funding for this research was provided by the MacArthur Foundation Global Studies Fellowship at the University of Wisconsin and the Social Science Research Council’s Program on the Corporation as a Social Institution.