Debate over the transfer and adaption of industrial models prompted by the globalization of Japanese manufacturers continues to draw attention. Most discussion has concerned issues of work and employment relations, where the new multinationals have interacted most directly with host societies. Two polar views have framed the debate. On the one hand, it has been argued that Japanese management systems cannot function effectively in the West ‐ or that if they can, the workforce suffers ‐ because they are culture‐bound, exploitative, and would be resisted by trades unions. Japanese practice has been interpreted as classic “Theory X” management. On the other hand, Japanese management systems are said to offer Western production workers new opportunities for teamwork, self‐expression and workplace democracy which should be grasped eagerly ‐ here Japan is interpreted as classic “Theory Y” management.

The characteristic Western “either/or” analytical approach may have set up a false dichotomy. Adler’s (1993) study of the Toyota/GM joint venture NUMMI, in California, revealed that rigid management structures were combined with opportunities for worker participation in the “learning bureaucracy”. Mair’s (1994a) study of Honda in Ohio showed how workers participated but in a way that was strictly channelled, and that the “single status” system both provided a degree of democracy and equality and permitted authority systems to function more effectively. Innovative management appeared to have reconciled the dichotomies that academic observers continued to debate.

Progress in understanding the Japanese multinationals has been further complicated by two other considerations. First, Japanese firms appear to have adapted their management systems to their new local environments in a process of “hybridization”. The questions become which elements of the model are retained and which are changed, and whether performance is inferior, equal, or even superior to that achieved in Japan (Boyer et al., 1998). Second, there are considerable variations in strategies, organizational forms and operations systems across Japanese companies (Freyssenet et al., 1998). Arguably, management systems at Nissan, for instance, share more in common with those at Ford than those at Toyota (an hypothesis for which Cusumano (1985) offers strong support, as do Western managers at UK automotive suppliers interviewed by the author who deal Toyota, Ford and Nissan on a daily basis). The idea of significant variations across Japanese firms has not yet been fully accepted, in part deflected by abstract theories such as “lean production” (Womack et al., 1990) which generalize from the Toyota case.

These theoretical issues are addressed in this case study of operations, organization and employment relations at the Honda factory at Swindon in southern England ‐ Honda of the UK Manufacturing (HUM). The HUM factory is the hub of Honda’s operations in the world’s largest market, and therefore of considerable strategic significance, although due to the relatively low European market shares of the Japanese producers HUM produced only about 6 per cent of Honda’s global vehicle output in 1998, with output anticipated to reach 150,000 per year by 1999 after six years of car production. Indeed the competitive environment in Europe has proved challenging for the Japanese, with southern European markets politically closed until the mid 1990s, fierce competition from domestic manufacturers of comparable products, marked exchange rate fluctuations, as well as political pressure for high “local content” even at low production volumes. Honda faced a particular challenge when rival BMW purchased long‐term local partner Rover Group in 1994.

The analytical approach adopted in this paper is based on the following proposition. The relationship between Honda’s global corporate strategy and the local operating environments of its factories is mediated by a nexus consisting of the production system, the organization of work, hiring and training processes, organizational processes and philosophies, and employment relations frameworks. This nexus defines Honda’s “industrial model” (Boyer et al., 1998; Freyssenet et al., 1998; Mair, 1998a).

Mair (1994a) offers a preliminary assessment based on research at the initial HUM engine plant. This paper incorporates further research at the car assembly plant. The first part examines the “received views” of organization at Honda in the management literature. The second part describes the operations management systems used at HUM, including the roles played by production workers. Third, HUM’s organizational processes and employment frameworks are analysed. The implications of the HUM case for the broader debates over “Japanization” are drawn out in the conclusions.

Empirical research included three plant visits to HUM and two further discussions with HUM managers, amounting to 12 interviews with managerial staff, all European: four in general management, three in purchasing, two in production management, two in human resource management, and one in materials handling. Honda is generally not very open to external researchers, and the conclusions of this paper are therefore provisional. However, the analysis of HUM has been aided by the author’s research at Honda in the USA, Canada, Belgium and Germany, including interviews with senior Japanese managers (sales, general, R&D) and product development engineers, as well as North American and European managers across the range of management functions.

Interpreting Honda in the management literature
Section:Previous sectionNext section
There is a “received view” of Honda in the management literature. In brief, Honda epitomizes the interpretation of Japanese practices as “Theory Y” management. Organization at Honda is said to be characterized by innovation, devolution of responsibility, and individualism, in stark contrast to the classical bureaucratic organization.

In new product development, for instance, emphasis is placed on the prominent role played by middle managers and junior engineers (Mito, 1990; Nonaka, 1988, 1991; Nonaka and Takeuchi, 1995). Engineers at Honda learn from experience; Peters (1988, pp. 316‐17) quotes Masaki Imai, author of Kaizen, to the effect that Honda has few PhDs on its staff. Garvin (1993, p. 80) cites Honda as a successful “learning organization”. This view parallels the “incrementalist” interpretation of Honda by strategy theorists (Mair, 1998b; Pascale, 1984).

Ohmae (1982, p. 220) also distinguishes Honda from the bureaucratic model, arguing that despite its size Honda has no organizational chart; indeed nobody externally knows how the company is organized except that it frequently employs project teams. Peters and Waterman (1982, p. 269) take their former McKinsey colleague’s view further; Honda does not have a structure, just innovative teams. At the same time Honda is said to have a “tight” organizational culture; “one of our favorite stories”, we are told approvingly (Peters and Waterman, 1982, p. 37), is that of the Honda worker on his way home who straightens the windshield wiper blades of all Honda cars because he cannot abide seeing flaws in them.

Absence of traditional hierarchy at Honda is stressed by Kanter (1989, p. 211), for whom the way Mr Honda worked directly with his young research engineers is a case of a high‐autonomy organization with a flat hierarchy. Similarly, Sampson (1995, p. 157) repeats a quotation from former company President Tadashi Kume (taken from Nonaka, 1988) on the integration of top and bottom by middle management, and argues that Honda gives many opportunities to junior staff.

Despite the strict environment, individualism is said to flourish, at least in certain respects. Ohmae (1982, p. 82) argues that Mr Honda was an obstinate persister in his plans, as does Sakiya (1987). Quinn (1991) follows the same line. Peters (1988, pp. 246‐7) approvingly quotes Pascale (1984) to the effect that Mr Honda was an “unreasonable” champion who succeeded. Whiteley (1991) tells stories drawn from Sakiya (1987), of a boss willing to get his hands dirty and to do what his workers did. Robbins (1991) describes how it was Mr Honda’s personal “passionate desire to succeed” that turned Honda into a multinational giant.

This general interpretation has also been applied to the work life and employment relations of Honda’s production line workers. The incrementalist strategy writer JB Quinn argues that Honda workers design their own production equipment; “workers had automated whole body‐panel and body‐assembly sections [unsupervised by engineers]” (1991, p. 293). Workers are supposed to have invented best‐selling Honda products. Quinn’s discussion of production organization focuses on equality, absence of seniority, willingness to “muck in’, and democracy; “on matters which affected them, associates [all employees] were asked their opinion” (1991, p. 293). This progressive interpretation of the production worker’s role is shared in Shook’s (1988) portrayal of Honda of America Manufacturing (HAM). But does it accurately depict operations, organization and employment relations at HUM?

Production operations at HUM
Section:Previous sectionNext section
The development of HUM has in part reflected Honda’s partnership with Rover Group, which has played a significant role in the company’s European entry strategy (Table I). The HUM site developed from inspecting Honda cars made by Rover to engine manufacture in support of Rover’s manufacture of Rover and Honda models before the HUM car assembly plant opened in 1993 and Rover production of Honda models was phased out.

HUM economics
Only the close collaboration with Rover Group made the relatively low volume HUM assembly plant economic. With versions of both HUM car models also made by Rover Group, many suppliers were shared (hence higher components volumes at suppliers), as were some product development costs. BMW’s 1994 purchase of Rover Group therefore posed questions about the viability of HUM, particularly since Honda’s market position in Europe did not permit output of the Accord and Civic to be raised sufficiently to generate economies of scale. Thus a “third model” (Civic estate) was added in 1998, with plans under way to add a fourth entirely different model which would raise output beyond 150,000 per year. HUM will become a further version of the Honda “flexifactory” concept (Mair, 1994b). Factory‐level economies of scale will be achieved through the accumulation of low‐volume models ‐ unlike HAM but similar to Japan. A wide range of engines is also made. Product development costs will be kept low since each new model (with the partial exception of the Civic estate) is spun off from a Japanese domestic market model. A remaining problem will be low volumes per component at European suppliers. Accordingly, HUM managers perceive the cost and flexibility of production tooling to be a key challenge in developing supplier performance.

Site layout
Honda Engineering (HE) has installed its newest production systems at each new Honda global investment, within an overall strategy which mixes experimentation and copying success from plant to plant. Hence HUM was closely modelled on the similar‐volume Honda of Canada Manufacturing (HCM) factory in Ontario, which first opened in 1986; many of the same Japanese production engineers set up both plants. The HCM factory had itself drawn lessons (although not the automated guided vehicles) from the experimental 1987 third assembly line at the Suzuka factory in Japan ‐ the principles of which were in turn partly inspired by the experimental Kumamoto motorcycle plant of the mid 1970s (Mito, 1990).

The first investment at the HUM assembly plant was a main assembly hall and paintshop. In one open space, the assembly hall contained a long U‐shaped assembly line, an adjacent body parts subassembly (welding) and body framing (welding) area as well as separate subassembly lines for the drivetrain (engine, gearbox and fittings), doors, and fascias. The sealed‐off paintshop was located at the end of the plant, linking the end of body framing to the start of the main assembly line. Small inventories of parts were stored at the opposite end, adjacent to the bend in the assembly line. Warehouses for receiving components from suppliers and batching them for assembly were opened elsewhere in Swindon. The testing of finished cars took place at the end of the assembly line, with an outdoor test track for all cars adjacent to this area. A separate stamping hall was built later, adjacent to the welding area within the main assembly hall.

Stamping
The Japanese‐built IHI tri‐axis automatic transfer press stamped major body panels. Production was at first restricted to a single shift, with output to expand as Honda brought more production in house as volumes rose and the relationship to Rover Group was wound down. A Japanese NKC conveyor system transferred stamped parts to the body framing station in welding.

Welding
All welding equipment was made by Honda Engineering in Japan, with a small HE maintenance team stationed at HUM. Small subassemblies were welded manually (although much of this work was outsourced, including to Rover Body and Pressings and other UK suppliers). Jigs in the subassembly area could be replaced to accommodate different models in seven minutes. The takt time was approximately 65 seconds. The welding area was very clean, and, as mentioned, was located only yards from the final assembly line.

The unique‐to‐Honda HE single station body framing cell (first developed in 1982 to make anticipated low volume production at HAM profitable), with several HE welding robots working within it, welded the major parts of the car bodies. Jigs for the framing station could be changed to accommodate a different model in seven minutes. On one visit by the author, only one car body awaited post‐welding rectification.

Paintshop
Joints in the welded bodies were sanded by hand prior to painting. A unique feature of the HUM paintshop was that car bodies on jigs turned 90° to proceed sideways and rolled over 360° in the pretreatment tanks to permit air pockets to escape and allow even applications of chemicals. Top‐coat paint spraying was highly automated (95 per cent painting by robot).

Goods inward
As the painted car bodies re‐entered the assembly hall parts were brought to the line by forklift from “goods inward”, which held stamped metal parts for body welding as well as components for final assembly. According to HUM managers, HUM held 0.3 days of parts (five hours) in the factory. Other stock was held at the Swindon warehouses: five days for parts from European suppliers and ten days for parts from Japan.

Main assembly line
The main assembly line was a single U shape, continuously moving. Doors were removed from the painted body at the start of the line and taken by overhead conveyor to the door sub‐assembly line, to facilitate work inside the car bodies and finishing the doors themselves. Bodies started the line facing forwards, as components like wiring and carpets were installed manually. For the second part of the line, around the U bend, bodies were turned 90° to advance sideways, facilitating reaching into bonnets and boots. They then faced forwards along the final length of the line. Workers stood on a rolling platform under the bodies so they did not have to walk while working.

Almost all assembly work was manual, sometimes incorporating hand held tools such as power screwdrivers. Aside from the moving line itself, two further types of automation were used. First, a number of mechanical aids held heavy components, such as an arm holding the sun‐roof subassembly which a production worker inserted through the front window as the part was fixed in place. Similar mechanical aids held the wheels and the battery while they were fitted manually and with power tools. The experiments HE had undertaken at the third Suzuka line with high levels of automation in assembly revealed that the combination of AGVs and assembly robots was not cost‐effective except in limited cases such as windscreen and power train installation. Since then, HE’s focus in assembly line automation had been to make manual work less physically arduous with these low‐technology aids. It was now argued that people performed some tasks better (at least in economic terms) than machines.

Second, HE automated cells installed front and rear windscreens (gluing and placing in position), as well as engine/gearbox/front suspension and the separate rear axle/suspension. At the latter cell, bodies were temporarily detached from the line and held immobile for machines to screw the mechanical components in place.

Instrument panels/fascias, engines/gearboxes/suspensions, and doors were sub‐assembled on small moving U‐shaped lines adjacent to the main line, synchronized with assembly of the exact car body into which they would be fitted. On the engine assembly line, engines stopped at each station (there was no moving platform under the line) for the worker to complete tasks before sending the engine on ‐ this “free‐flow” line was one of the innovations at the Kumamoto factory. Overhead conveyors delivered subassembled fascias and doors to the main assembly line for direct assembly into the correct body.

Batching and logistics
Honda utilizes a small batch production system without kanbans, quite distinct from the one‐piece‐flow systems often associated with Japan. This system could be observed at HUM. Accord and Civic models were each made for a half‐shift, with jigs changed during rest breaks and thus a maximum of 3‐4 jig changes per day. Cars were further batched into sets of 60 or 30 cars precisely the same except sometimes for exterior colour where there might be two 15‐car sets within a batch of 30 (same interior colour). One batch, for instance, might consist of 30 Accord models each of exactly the same specification destined for the German market. The next batch might be 60 cars, each identical, for the UK market.

At goods‐inward and lineside, components were packed into specially made returnable plastic boxes or racks of 60 or multiples and factors of 60 ‐ for instance, lots of six for fuel tanks, lots of 240 or 360 for small pressings. These parts had been made by suppliers specifically for the actual batch of cars in which they would be used. One advantage of the small batch system was its function as an automatic control system for quantities supplied by suppliers and parts properly installed by HUM workers. HUM managers explained that the company did not make “smarties” (reference to the multicoloured sweets) since this would increase the likelihood of errors.

The system required parts quality to be high. Very small stocks of spare parts were kept at lineside for use in emergency (poor quality part, damage during assembly). Quality problems in, for instance, the paintshop, might require one or two car bodies to be inserted between later batches to bring the system back into equilibrium.

In contrast to this seemingly rigid production system, HUM exhibited the same “structural” flexibility as the other Honda “flexifactories”. Rather than plan and build a single “200,000 units per year” factory initially, Honda Engineering is able to make strategic investments step by step, as indicated by the incremental ramp‐up of production at HUM, or the later investment in a stamping facility. A further illustration concerned the engine plant. The original plan had been for 70,000 1.6 litre engines per year to supply Rover’s Longbridge factory. The success of the new Rover models exceeded expectations, and in early 1992 Rover asked Honda to increase engine supply. Honda agreed to raise production to a rate of 116,000 per year later that same year, and HUM met the target.

Production work, employee flexibility and involvement
Section:Previous sectionNext section
What did Honda’s production system imply for the organization of work at HUM? This question can be broken down into three parts. First, how did HUM itself describe the roles of production workers? Second, what was the relationship between the production system and employee tasks, including flexibility? Third, what was the relationship between the production system and employee involvement in improving the performance of the production system? These questions bring us closer to the theoretical issues raised at the start of the paper.

The HUM philosophy
The HUM motto was “Quality through teamwork and technology”. Technology has been discussed above. The “teamwork” concept appeared to refer to a broad emphasis on giving employees the most responsibility possible within the constraints of the production system, combined with a broad concept of co‐operation among employees (rather than to describe a particular model of work organization).

Discipline and flexibility
Daily production activities were set within a basic 37 hour working week (4 × 7.5 hours and 4.5 hours on Fridays plus 5 × 30 minute breaks). Start times and break times were adhered to strictly. All employees were obliged to wear company overalls and identical baseball‐type hats. Daily work for the vast majority of production employees involved manual work following standard operating procedures for each group of tasks fitted into an approximately one minute takt time. Task variation was reduced ‐ compared to single‐piece‐flow systems ‐ by small batch production. Work at every assembly station was precisely the same for at least half an hour ‐ the same at most stations for half a shift. The mid‐shift change from Accord to Civic was explained in part as making production work a little more varied.

According to a HUM personnel manager, not all new employees (see below on hiring practices) coped well with the rigours of the system. There was an elevated level of turnover during the first few weeks of employment, which Honda accepted, due to the “regimented operation of the factory”.

Work assignments were otherwise varied or flexible in four ways. First, there was rotation within the production teams of 4‐20 workers (size a function of the technical organization of production). Training matrices indicated which tasks each employee had learned. Workers were rotated by their team leader. The periodicity of rotation appeared to vary, up to daily rotation in some cases. There was no rotation between teams, which HUM managers argued might be good in principle ‐ largely to provide variety of work ‐ but would disrupt the stable work patterns required for quality. It was also argued that HUM was a new operation and basic skills still needed to be built up within teams.

The second form of flexibility was related to line balancing for car batches of different vehicle complexity. This affected only a few segments of production where more work was required on certain batches ‐ and Honda further minimized the “line balancing problem” by the market offer of options packages rather than “customized” cars. Manning levels at affected stations were set for “low spec” cars. Team leaders and supervisors then planned their tasks ‐ usually possible days in advance ‐ so that they could work on the line as each batch of 30 or 60 high spec cars passed through.

The third form of flexibility brought employees from other areas of the factory, notably the offices, to the production line. This might be to free up a production worker or team leader whose knowledge was required to solve an urgent problem, or in case of illness. Each employee whose main job was administration was trained and ready to undertake a production task when required. On the day of one visit by the author, a purchasing manager had spent part of the day bolting on manifold covers in the engine plant due to a problem there.

The final form of flexibility was the infrequent “structural flexibility” which required selected workers to move into new production areas as they were constructed. The HUM strategy was to assign experienced workers to the new areas and “infill” behind them with new employees. Hence the car assembly plant was initially staffed with workers from the engine plant.

In general, employees were expected to be flexible to meet the requirements of the production system. However, while work variation was seen as beneficial in reducing boredom, it did not seem to be viewed as beneficial to the production system itself (for instance, by stimulating worker interest).

Employee involvement
In the initial years of HUM operations, plant management was keen to stress (directly to the author) the high levels of “ownership” HUM workers had over their work areas, including its technology. Workers were said to have set up whole machining areas themselves after observing the systems in Japan, and to have improved the production process by recycling waste (see Mair, 1994a, for further examples). This emphasis fitted well with the interpretation offered by Quinn (1991) and Shook (1988).

Yet the attractive vision of employee empowerment at HUM must be treated with some caution. In the first place, most examples of employee involvement at Honda in Japan (Economist Intelligence Unit source), at HAM (Shook, 1988), and at HUM (author research) referred to the margins of the main production process (ranging from car parking and restaurant facilities to the recycling of waste materials). This bias emphasizes that the main production processes, including the standard operating procedures, have been established by highly skilled HE production engineers. Indeed the standard operating procedures which will be used in the factories are built into the product design process at Honda. While new processes may sometimes be shown to production employees for verification, Honda Engineering has developed such a high level of expertise in systems which have been iteratively modified over generations of Honda products that there is usually little scope for shopfloor improvement.

No suggestions systems or quality circles were set up in HUM’s early years. Management argued that involvement was “part of the job”, a means to improve job security, rather than rewarded separately. Quality circles (called by Honda “NH circles”) were started in 1996 after seven years of operations. It was explained that workers first needed to understand the system if they were to channel their ideas within its parameters. “Got a theme, form a team” was adopted as the HUM slogan. By 1996 all employees were being encouraged to learn basic statistical and other management tools, including “pareto, characteristics diagram, histogram, stratification, scatter diagram, and check sheet”. By 1996 too, “associates committees” had been set up across the plant to receive employee suggestions dropped in boxes.

Employee selection, training and development
Section:Previous sectionNext section
Employee selection
The first decision to affect the selection of production workers was the choice of the Swindon site. Swindon is a substantial town in an area of low unemployment dominated by service sector industries, about one‐and‐half hours south of the heart of the West Midlands automotive industry. It is also home to a major Rover Group factory (Rover Body and Pressings: body stamping plant) and was previously the location of railway engineering workshops which had an active unionized workforce. The most obvious advantages of the location were a large site adjacent to the town and proximity to Heathrow Airport, one hour away.

Honda and Rover Group adhered to an “unwritten rule” not to accept applicants already employed at the other. Honda gave applicants with potential two interviews, each with two interviewers present. There was no extensive testing of applicants, since Honda sought “good basic people who can grow with Honda”, rather than “stars who can land with their feet running”. General aptitude and attitude, job histories and previous records were considered more significant than formal qualifications.

There was a considerable Japanese presence among the highly skilled engineers and managers, though numbers and precise roles were difficult to ascertain. HUM management was keen to stress the level of local control. Thus while the HUM managing director was Japanese, he was said to work at the Honda Europe headquarters and rarely visit Swindon. According to British managers, the Japanese associates “help us”. It was the British managers who identified what in the HUM organization should reflect the local environment and what should reflect Honda practices elsewhere in the world. While the British managers may have exaggerated their roles somewhat, Honda had indeed permitted them significant leeway in developing appropriate organization systems, especially in the areas of hiring, training, employee terms and conditions, as well as management training and development.

On the other hand, the production system and Honda standards were not negotiable. The limited available information suggested that there were two parallel managerial hierarchies ‐ as at HAM (Mair, 1994a). British employees managed the plant in general, human resources, purchasing, and the direct hierarchy above the factory floor. Highly skilled product and production engineers were mostly Japanese ‐ with their deep knowledge of the Honda production system. The Japanese staff were assigned to HUM for periods which varied from a few months to several years. These staff might be reassigned on a global basis at short notice, with or without their families. According to senior Japanese managers interviewed, ambitious Western managers were expected to accept the same rules as the Japanese, since the reassignment of key staff to new projects, wherever located, was required to give Honda strategic agility.

A group of young British manufacturing engineers was also hired, probably so that some of the Japanese could later be replaced. Of 30 recruited during one period, only two had a motor industry background. Moreover, there was no scheme to recruit graduates directly from university for a “graduate training” programme. Indeed graduates were not held to have particular advantages. More important were individual capacities and a “broad mind”. The British engineers had been hired primarily to learn and operate an existing and highly specified system, not to contribute (yet, at least) or to question.

Training and development
Training for production workers was mostly “on‐the‐job”. Training matrices in each work area matched employees against a series of tasks. Experienced workers were able to do all the jobs in their team ‐ recent hires perhaps only two or three. A group of British managers referred knowingly to “what is lovingly called on‐the‐job‐training”, but declined to interpret the adjectival phrase. (The author believes that it referred to the way new recruits were assigned to front‐line jobs immediately with training squarely focused on undertaking more production tasks.) HUM made no use of the British “National Vocational Qualifications” (NVQ) system. There was no apprenticeship system for training skilled (e.g. maintenance) workers. No financial assistance was available to employees seeking external qualifications. In short, training focused on particular jobs, through what was described as a “natural” learning process. An internal study group set up in 1996 to ascertain how HUM could act as a “good corporate citizen” by co‐operating with local training and education institutions found it hard to work out how Honda might benefit in return.

Both graduates and skilled workers were initially assigned for three‐six months to work as production workers. There were pragmatic benefits from providing them with first‐hand experience of production work and workers. There was also a social/behavioural goal (see below). Several employees at various levels had received training in Japan, and senior British managers had studied Honda plants elsewhere.

Most managerial positions, especially in the hierarchy of control over production, were filled through internal promotions. All production co‐ordinators (the level above team leader), and several section leaders and departmental managers had first been hired as production workers. This policy reflected the importance of managers knowing the Honda production system and Honda knowing the candidate managers. For instance, to be a good buyer (supplier liaison in the purchasing department) it helped to have worked on the line and know the problems of trying to fit poorly made parts. The absence of an apprenticeship system was explained in terms of the absence of a barrier to promotion for production workers. At HUM, all employees could “grow”. Twice per year an appraisal process selects good performers for promotion. Those judged to have the right aptitude and attitude may be moved horizontally as well as vertically, to improve their understanding of the Honda system.

Organizational philosophies and processes
Section:Previous sectionNext section
An examination of organizational philosophies and processes rounds out this analysis of the transfer and adaption of management methods at HUM. HUM’s organization may not be easy for new employees to grasp initially, according to HUM managers. For the first six months, the minds of new employees may stay closed, before they open up in realization of how the organization works ‐ Shook (1988) made a similar comment about HAM.

From the viewpoint of plant management, the components of the organizational philosophy form a pyramid ‐ “the rock of good team work”:

1. quality products are built on a basis of:
2. good team work: ‐ this in turn needs solid foundations:
3. single status and job flexibility, which drive out, respectively, class differentiation and job demarcation;
4. in turn, good communications, participation, competitive working conditions and fairness undergird these; and
5. finally, at base, there is a combination of togetherness and self‐development.
The principal components of the above framework can usefully be examined in more detail.

The single status system
Single status is manifest in the sharing of a single shared pension scheme, sick pay rules, holiday rules, clothing, car parking, canteens and changing rooms. It is also apparent in open‐plan offices. But single status was expected to be more than superficial; it is a powerful system at the heart of organization at HUM, and played at least four significant roles.

Permits job flexibility. Single status is firmly linked to a structural and short‐term job flexibility straddling traditional British “demarcations” of status and skills. Job demarcation is described as “living in mental prisons” and “asking for redundancy”. Not only are there no job descriptions, but no secretaries are employed at HUM, with employees preparing their own papers when necessary. No minutes are taken of meetings (“we don’t need that rubbish with servants”). Instead, participants take photographs of the white‐boards upon which agreed decisions are written during meetings.

Promotes creativity. Single status was said to permit new projects to be driven from the factory floor, releasing the creative power of all employees. Because single status means equal opportunities, it encourages employees to “graft and grow”, to achieve, with no barriers to recognition.

Role of single status in the UK context. A great deal was made at HUM of the importance of dispensing with the “class” and “status” attitudes which have traditionally prevailed in British industry and acted as barriers to flexibility, mobility and communication. In this context single status fitted tightly with what can best be described as HUM’s ideological offensive against status‐oriented attitudes, and was further supported by the cross‐training of employees in each others’ tasks.

All Honda employees were called “associates” ‐ as at HAM ‐ to indicate a sense of belonging and to “get away from the class laden rubbish” of distinguishing staff (white collar) from hourly‐paid (blue collar) employees. It was said to be “abhorrent” to allocate employees to different canteens according to “caste”. Indeed the framing and language of the ideological offensive appeared particularly British, reflecting a very British working class antipathy towards status‐conscious managerial attitudes, but which was now being broadcast by management itself. Hence Honda did not want “lounge lizards in suits”, or people who felt superior because of their clothing, where they were educated. Part of the reason university graduates were assigned to production work was to “knock that status stuff out of them”. Status differences were said to be “rubbish from the past that has to be cut out”. HUM managers claimed to “come down hard” on employees deemed to be “playing politics”. New employees were told that divisions and barriers are not tolerated. So strongly was this new culture communicated, that the plant manager, who insisted that not only Honda employees but all visitors to the factory abide by the uniform rules and wear their baseball caps was nicknamed “Jones the Hat” by Rover Group managers interviewed by the author.

Clarifies vertical hierarchy. The single status system did not prevent the functioning of managerial hierarchy. Indeed the opposite is true, since it cleared away a number of impediments to authority structures and communication. Just as the Honda production system was being protected from the outside interference of “inappropriate” new ideas, so too lines of command were clarified by the removal of potential obstacles.

Supports teamwork. Finally, single status removed impediments to communication and collaboration in small groups, enabling more fruitful discussions and more effective decisions.

Communication, participation and fairness
HUM managers argued that open communication was a hallmark of HUM’s organization. A further reason for the training of office staff in production work was to break down potential communication barriers. Managers held meetings with randomly selected groups of employees to discuss their concerns. All employees were consulted when the plant moved from a 39‐hour to a 37‐hour working week, with many small group discussions held. The relatively young workforce decided to prioritize leaving work early on Fridays, and so the second shift was brought forward to start immediately after the shortened first shift so that second‐shift workers, who normally finished at midnight, could leave at the end of the afternoon. As with suggestions and quality circles, this type of exercise in broad participation had no impact on the effective functioning of the Honda production system.

In 1991 HUM adopted a disciplinary system partly based on that at HAM (Shook, 1988). Employees facing dismissal could now choose to appeal either to the HUM managing director or to a randomly selected committee of seven production associate peers. In 1993 the committee of peers was chosen for the first time.

Competitive working conditions
Basic working conditions, in terms of level of pay and benefits, were established by reference to local and British labour market norms. However, in keeping with single status, all production operators, including assembly and welding workers, were paid the same salary after a graduated increase during their first year of employment. A bonus equivalent to about 5 per cent of salary was paid each month to those with full attendance. There are also payments to recognize shiftwork and overtime work.

As discussed above, HUM did not make individual cash payments to employees who had made particular contributions. Instead, they were given vouchers for restaurant meals and clothing. Similarly, a good safety record was rewarded by a company donation to local charities.

The 25 days of paid holiday per year was in line with UK norms. Private health‐care insurance for all employees, while not the norm in the UK, was explained in terms of ensuring that no employees were prevented from returning to work after illness or injury due to the sometimes slow response of the National Health Service. Managers also received a “company car” as part of their benefit package, in line with common British practice. HUM managers believed that this adaptation to the UK environment made them unique at Honda globally, although the author’s research in North America suggests that this is not the case.

Honda’s strategy towards trade union representation was clearly antagonistic, and the HUM plant appeared to be the first automobile assembly plant in Europe without a trade union presence for several decades. There was no real campaign for union recognition beyond an initial sparring in which local union representatives (interviewed by the author) exhibited the same lack of understanding of Honda’s production system and organization as their US counterparts a decade before, anticipating that the workforce would soon become disillusioned with dirty working conditions and poor management (Mair, 1994a). Accordingly, no campaign appropriate to HUM workers was developed.

HUM managers argued that there was no need for trade union representation or a works council, due to the “excellent communications system”, primarily composed of a myriad of “associates committees” in each work area. Employees were told that managers would feel a “personal sense of failure” if the workforce sought trade union representation.

Conclusions
Section:Previous sectionNext section
This paper has investigated operations, employment relations and organization at Honda of the UK Manufacturing, a nexus which forms one of the key “bridges” between Honda as a multinational enterprise and its local operating environment (Mair, 1997). The empirical findings can now be linked to the theoretical questions introduced at the start of the paper: the interpretation of work and employment at Japanese manufacturers, the relationship between transfer and adaption of management techniques, and our understanding of Honda as distinct from a general view of Japan.

The preponderance of evidence from HUM in the 1990s points towards a “Theory X” interpretation of the production system and organization of work, which were found to be very tightly controlled, with little variety of work or scope for individual initiative. Honda had sought to hire production workers to fit into an existing system rather than question it. External qualifications (and therefore knowledge) were not viewed as relevant. Significantly, this finding also encompassed the British engineers who had been hired to eventually replace some of the Japanese technical staff.

How should this evidence, which seems to flatly contradict the received view of Honda in the management literature, be interpreted? There are a number of possible explanations, each of which appears to offer a partial truth, but which together may offer a fuller contextual understanding of HUM.

First, HUM may be interpreted as at the start of its own development trajectory, in a phase focusing on the learning of a very specific Honda production system. Since this system is complex and subtle, a thorough understanding ‐ and therefore a prolonged period of learning ‐ is required before employees can make “independent” decisions which fit within its bounds. This may explain the absence of interest in external qualifications and the seven year lag before establishment of a quality circle scheme.

Second, while the preponderance of evidence at HUM may lean towards a “Theory X” interpretation, weighing the balance of evidence in this way may not be the most appropriate evaluative technique, since it remains within a dichotomous world view. The significant issue may be how Honda combines a “tight” production system with the “looseness” of individual innovation. It seems likely that different functions at Honda have a different balance of control and creativity. Mito (1990) argues that the organization of product research at Honda R&D reflects a creative and risky process, while product development is tightly controlled to reduce risk. From this perspective, manufacturing at Honda is a function which is tightly controlled to produce high quality and high efficiency, with little scope for individual contribution; the received view of Honda may be a generalization based on observations of other functions.

Another possibility is that there is indeed scope for input from HUM production workers, or rather that there will be once the production system is thoroughly embedded, with a “tack” towards more creativity at a later phase. In this case we could anticipate geographical variations at Honda plants related to their development trajectories. A function performed in Japan may have a certain organizational openness, which is possible due to experience of employees, in comparison to the same function performed in a new transplant in the West where Honda systems have not yet bedded down. This may partly explain the insistence of plant management that individual contribution is valued at HUM; it is valued in principle, and will be more valued in practice in future. And yet creativity seems likely to be strictly confined within the principles of the Honda production system. Thus production at HUM offers one functional, geographical and temporal window onto the combination of control and creativity at Honda.

To what extent has the Honda model been transferred and/or adapted? The evidence at HUM confirms Mair’s (1997) overall assessment of Honda’s global‐local multinational form. From a technical and organizational viewpoint the production system is being transferred to other countries entirely intact, while adaption is confined to “softer” aspects of organization and employment relations to produce a fit between the production system and the local operating environment. The question is not whether there has been transfer or adaption, but how adaption has taken place to permit (been combined with) transfer. This explains why the creative freedom of local managers focuses on the interface between the production system and the labour market. Key elements of the framework ‐ single status in particular ‐ have been learned from studying experience at HAM, while the ideological framework in which status and equality are discussed at HUM appears peculiarly British.

Finally, what if anything does this analysis of HUM tell us about “Japanization”? A measure of caution may be appropriate. From the author’s perspective, this study is best viewed as one contribution to the available knowledge on Honda and the internationalization of Japanese manufacturers. It cannot be assumed to represent the class of “Japanese manufacturers”. Only systematic in‐depth comparative case study of different Japanese firms can answer the question of how representative one case study can be. Indeed a key conclusion of this analysis is that manufacturing at HUM during the 1990s cannot even be taken as “representative” of Honda ‐ let alone “Japanization”.