Work organization in lean production and traditional plants: what are the differences?
Most publications on lean production (LP) devote some space to the question of the importance of human resources. It is only rarely that work organization practices (WOP) are clearly defined and the specific role of each of them within the LP model is dealt with. Generally, the operations management literature associates WOP to LP through examples with the implicit idea that all the WOP quoted are necessary and also typical of LP. This study calls for a more critical view on WOP in relation to LP: its aim is to contribute to the construction of a grounded theory on work organization practices in lean production.
This study first approaches the problem by comparing lean production with Fordism and Taylorism. It then identifies the work organization practices that are associated with lean production in the literature and examines their supporting role in relation to lean production practices. The result of this analysis is inserted in a reference framework for the study of the association existing between work organization and lean production practices. This framework is also used to develop hypotheses which are subsequently tested using data collected in a multi-purpose study on lean production.
Lean production versus Taylorism and Fordism
In the 1980s a research group from MIT conceptualized the Japanese model - of which Toyotism seemed to present the purest version - as "lean production". These authors, and others also, affirm the universal validity of this model, notwithstanding its Japanese origin. Many scholars see lean production as a complex of organizational innovations which has the same strategic importance as that of Taylorism and Fordism but which has since undergone a change of point of view in comparison with them[1-3].
The terms "lean production" or minimum workshop", as Ohno says, are inspired by the fact that, compared with Fordism, the lean model requires less stock, less space, less movement of material, less time to set up the machinery, a smaller workforce, fewer computer systems and more frugal technology. As well as responding to the need to be cost effective, this characteristic also constitutes a general principle that inspires a philosophy of essentiality and which makes every superfluous element seem wasteful.
Several authors have asked themselves whether, and to what extent, the lean production system does actually represent a production revolution compared with the previous "Fordist-Taylorist" model, and to what extent, on the other hand, it just constitutes a mere adjustment to changed market conditions[2,5-7].
The lean production system demonstrates a certain continuity both as regards the Taylorist organizational model insofar as both are seeking to obtain total productivity, and as regards the Fordist production model as far as the desire to obtain synchronicity of the production flow is concerned.
In fact, lean production also places great importance on maximizing productivity. It does, however, transfer the Tayloristic workforce organizational principle, which aims at the total productivity of each individual member, to the organization of the whole company. There is a development therefore, from one optimal way to complete every individual operation to a "one best way" systemic approach which can be applied to operations in the whole factory, and further still, to the whole system which is involved in the production of a particular commodity.
The lean production system manages to integrate a complex plurality of productive segments into one single synchronic flow (take, for example, the pull system within the plant and the pull link with the market and with suppliers). This corresponds to Henry Ford's production dream of the synchronic factory, or indeed to the idea of the production space as an integrated unit linked together in one perfectly continuous cycle (see the Ford project planned in the 1920s in River Rouge at Dearborn which failed because of the inability to manage the production flows of differentiated products in a synchronic way). Nevertheless, the lean production system can be differentiated from the Taylorist-Fordist system on account of some basic characteristics which are such that the lean production system can be deemed a new production philosophy.
In the first place, it differentiates itself through its relationship with the market. In fact, the market dominates production and the lean production system adapts itself to this new condition through communications which move upstream, and through forward movements of the production flow pulled by downstream needs which originate from actual market requirements, compared with the traditional push system in which production is pushed downstream in line with pre-determined plans. The second deep-seated difference concerns the workforce. Human resources are no longer seen as a resource which naturally tries to resist the supply of work, but as a resource which naturally wants to work and indeed has to give more as regards the supply of work. Thus, the dualistic vision of factory society has faded, giving way to the idea of a unified community with a strong cultural identity, and the hostile climate embedded in Taylorism has therefore been reversed towards a climate of greater collaboration.
Many scholars have wondered, however, if human labour in lean production does actually become more intelligent compared with Taylorism/Fordism, or whether it only becomes more demanding, or indeed whether both phenomena occur simultaneously. This ambiguity is the result of the fact that a production system which culminates in just-in-time (JIT) is very efficient if everything runs perfectly, but is extremely fragile if there is any kind of problem whatsoever, because with JIT the company gives up the expensive security supplied by excess resources and relies on the synchronization of its various departments. Consequently, in order to be able to function in a lean system, all the resources being used in the production process have to be foreseeable and reliable and above all there has to be a totally collaborative social environment. But, as Oliver and Wilkinson say, this requirement can be ambivalent. On the one hand, JIT requires the workers to be totally subservient to production needs, while on the other, the success of the company's adoption of JIT depends totally on the willingness of workers to collaborate.
Taylorism places a great deal of emphasis on close control of the jobs to be done. The ex ante control is realized through such principles as maximum job fragmentation, the divorce of planning and doing, the divorce of direct and indirect labour, the minimization of skill requirements and job learning time. Through the classical assembly line principle, Fordism also pursues the same aim by means of the technology itself, and not in this case through a series of instructions. In the lean model, however, job responsibilities do not have precise delimitations, and employees are encouraged to participate in the making of decisions concerning production. This favours the learning of new skills and also the taking on of responsibilities which, formally speaking, had not been assigned at the beginning.
In the Taylorist/Fordist system the ex post control system relies heavily on hierarchy. But the economies of extended division of labour tend to be offset by the diseconomies due to dissatisfaction, management control structures and cooperation costs. The principles of Taylorism and Fordism therefore reveal a basic paradox "that the tighter the control of labour power, the more control is needed"[9,10]. But in any case, since the control was purely disciplinary and extrinsic, it ended up being inefficient and the workers managed to create shadowy areas in which with tacit astuteness they adapted the work to their own working rhythms. In lean production, however, the JIT and kaizen mechanisms are incorporated into the process and have a retroactive effect on work so that from one point of view it reacts immediately to the needs of production flow and from the other it becomes transparent, through forms of management by sight and the progressive reduction of stocks. In this way, the workers are structurally stimulated to move outside the bureaucratic confines of particular job responsibilities. At the same time, the dimension of increasing responsibility dissociates itself from the dimension of diminishing autonomy and dissolves in interiorized conformity with systemic needs. The structural pressure to move towards this transparent reactivity is combined with a cultural pressure which is aimed at obtaining the greatest employee willingness[5,6].
In order to eliminate unnecessary sequences and movements, in lean production the Tayloristic principle of the one best way is replaced by the principle of stock reduction. While the one best way method bureaucratically imposes work spaces, times and actions, the reduction of stock becomes embedded within the production process, thus producing full visibility as regards the work process, and the involvement of the workers' intelligence in the elimination of redundant actions and idle time. In this way, the one best way method is never definitive or imposed from above, but is rather an idea with a mobile limit, which can be extended through continuous improvement obtained by successive innovations, each of which appears in its turn to be superable and open to further innovation. In the same way, the standardization of procedures is no longer defined by a restricted group of experts and imposed hierarchically on the workforce, but is realized with the active involvement of the whole workforce, who indeed believe that the standardization of procedures can always be improved[5,6].
The lean model undoubtedly surpasses Fordism, both from the production point of view, since it becomes flexible and gains in quality, and from the human point of view, where worker involvement takes the place of bureaucratic controls. But this does not mean that work becomes freer or more relaxing from an individual point of view; on the contrary, the structural constraints and social pressure lead to working rhythms which are ever more frenetic. An equally complex evaluation can be made of Taylorism. There is no longer the impoverishment of different job duties nor hierarchical impositions, but the constraints imposed by standardization and the reduction of work time increase, since it is the workers themselves who pool their ideas in order to find the best solutions and who are then obliged to follow their own suggestions in the uninterrupted spiral of continuous improvement.
Human resource management and lean production practices
Many articles concerning lean production contain some paragraphs about the importance of training workers, and also comment on the importance of worker and management commitment, but then focus on technical aspects[12-141. In addition, there is no consensus regarding the extent to which, and the way in which, the adoption of lean production has implications for human resources. Some scholars emphasize how there is increased autonomy and job variety, while others underline the increase in work intensity and stress.
Groebner and Merz identified the following as the most significant changes that JIT creates in the work environment: performance evaluation, group interaction, job flexibility, mode of supervision and potential loss of security. Other major ways in which JIT appears to affect the people on the shop-floor are the loss of individual autonomy, the loss of team autonomy, and the loss of freedom to take risks[15,17,18]. It might be useful at this point to consider human resource management in Japanese companies in the UK and the USA. In fact, as Wilkinson and Oliver write:
Japan has abandoned extensive welfare provision, lifetime employment and seniority wages and promotion. Japanese manufacturers have brought about total quality and just-in-time manufacturing systems and the team concept which goes with them. On both sides of the Atlantic, Japanese Human Resource Management policies entail careful attention to selection, training, consultation and communication, single status facilities, harmonized terms and conditions, the minimization of job classes, and individual appraisals emphasizing cooperation, teamwork and flexibility. Such practices underwrite an efficient manufacturing system dependent on teamwork, strict discipline, flexibility and attention to detail.
Work organization in lean production deals with more than job design. It emphasizes management style, skill and values and aims to incorporate job design into the organization's employment strategy. The approach to job design is therefore characterized by co-operativeness, group problem solving, and attitude control in what can be described as the social organization of work. Nevertheless, this social organization of work system constitutes a sophisticated control system of employee behaviour[19,20].
Employee involvement is perhaps the aspect on which the various scholars working in this field agree most. This involvement manifests itself in many forms but in particular through:
* the right-duty of the workers to interrupt production flow whenever they notice anomalies or defects;
* the exchange of positions within the work group and the habit of "giving each other a hand" in moments of difficulty;
* the adaptation of work teams to variations in job duties and in the production flow;
* the commitment of each worker to the continuous improvement of each production factor.
Both JIT and total quality management (TQM) rely on team concept. In a JIT system a worker cannot produce another unit until the worker at the next station signals that another unit is needed. On the other hand a difficulty at the upstream station could block work being carried out downstream. The output of each worker is therefore strictly linked to the output (both in terms of volume and quality) of the other workers in the section. All workers must act as a team, then, rather than as individuals and performance must be evaluated in terms of the output of the entire group rather than that of the individual. Teamwork and group problem solving allow decision making to be decentralized and therefore variance and uncertainty is managed more easily. It also makes it possible to achieve a better "ownership" of the process on the part of the workers, and this, in its turn, is at the root of better improvement suggestions. Employee involvement is also shown through the contribution of ideas towards the improvement of the production system. In order to push the employee in this direction, it is necessary for the supervisor and managers to motivate him/her, however, and that true interest is shown in the suggestions he/she makes.
The solution of problems by workers places more power on workers. Supervisors, therefore, have to give up a certain amount of control, and have to change their roles and their management style, acting more as facilitators, co-ordinators and trainers. Team organization typically places the supervisor in a much more demanding role than he has traditionally held. The practice of problem solving also includes the participation of production engineers. In fact, there is not a real division of labour between workers and engineers in finding solutions to problems, and there is concrete recognition of the fact that they are working together in limited salary differences and that career possibilities in the long term are also extended to workers. Opportunities of meetings with managers and technicians are frequent in a system in which the exchange of information, both informal, through frequent and direct communication, and formal in occasions such as continuous improvement groups, team briefings and quality circles are considered to be very important.
The feedback to employees thus provides them with a means of learning and maintaining task-oriented behaviours[24,25]. It is, therefore, important for there to be the selection of appropriate feedback during continuous improvement activities. Different types of feedback have to be given at different organizational levels. Process control requires continuous data collection as regards process behaviour. It has been shown that when feedback comes back from the individual operator to himself (short, fast loops), better quality performance is reached[26,27]. Internal information on quality gathered by product inspection and internal process auditing constitutes a big part of the feedback. It is important to present this information immediately in order to facilitate research on causes and to intervene rapidly with corrective action.
Lean production privileges process rather than functionally-based management. At shopfloor level this means that horizontal co-ordination prevails over hierarchical co-ordination. In order to work properly, horizontal co-ordination needs each department to identify and tackle its own internal problems independently(29]:
Under JIT, employees become more self managing than in a command-and-control factory They investigate process improvements and monitor quality themselves; they consequently enjoy immediate feedback regarding their own performance. (Managers don't need to tell them how they're doing. They help design the system and it tells them how they're doing).
At the same time with JIT, buffers are reduced and therefore also slack and idle time. As a result, employees have less time, if any, for vertical tasks (i.e. administrative or other duties traditionally performed by supervisors), to take short breaks, or to assemble for a team meeting. In its pure form then, JIT could well lead to higher work intensity and stress levels among line operators. The same argument extends up to work teams. Without stocks to "de-couple" a work unit from other manufacturing processes, there is little scope for the team to exercise its discretion over what activities to perform or when to perform them. In addition to this, there are cases in which the adoption of daily instead of monthly targets restricts the freedom to make changes to the manufacturing process on a trial basis. Public displays of individual and group performance have great control power. In some realizations of this system, self-regulating teams have created an attendance system which relies on peer pressure or clan control to discourage tardiness, absenteeism and errors. The transgression or mistake of one team member creates problems for all team members. When one team member is absent or fails, his/her team mates will have to work that much harder to make up the slack.
Through increasing workers' sense of responsibility, the lean production system tries to find a remedy to the common practice of never stopping the production line, even when there is clear evidence of processing errors, thus putting off corrective action until a later time in the production cycle, and relinquishing the possibility of correcting the causes of the defects. Through autonomation (stop the line), retroaction and error autocorrection are immediate.
Because the goal of JIT is to produce to demand rather than to achieve smooth production runs, each worker must be cross-trained to perform several tasks so that he or she can step in where needed. In JIT, therefore, idle time generated by rigidities in traditional job descriptions must be eliminated. JIT workers are expected to be multi-skilled operators, who have the ability to run multiple machines, do their own quality control, solve quality problems, and do a variety of jobs involving a variety of skills and talents[17,33]. The aim is to reduce work-in-progress and to minimize any idle equipment, facilities or indirect staff. In addition, the use of statistical process control by employees documents and improves their own performance[22,26,27]. Low set-up times and frugality in the machinery (designed to reduce difficulties of use to the minimum) make it easier to overcome the traditional distinction between workers trained for the setting up of the machines and those trained in production.
In lean production good process management is very important. The precise definition and documentation of process management procedures and the setting up of procedures for cleanliness and organization of the workplace is also necessary. This reduces the likelihood of process disruption and operator error. This shopfloor documentation is used to facilitate change rather than to prescribe people's behaviour rigidly. It is used to facilitate improvements in methods - which means that the manuals do not gather dust on the shelf, but are constantly changed as improvements are made. These aspects of work organization in lean production seem amply to incorporate the principles of good job design indicated by Litter and Salaman (closure, incorporation of control and monitoring tasks, task variety, self regulation of the work speed, possibility of social interaction and co-operation among workers), and the job characteristics identified in the job enrichment model by Hackman and Oldham (skill variety, task identity, task significance, autonomy and feedback). Nevertheless, these work organization practices seem to be even more closely related to the twin track model of job redesign, which assumes that job satisfaction determiners and job performance are analytically distinct
In order to analyse the differences between traditional plants and lean production plants a theoretical reference framework is needed to clarify the characteristics of lean production and work organization practices. The various scholars working in the field agree that JIT and TQM are characterizing elements at the base of the lean production model. As far as work organization practices are concerned, this study will focus on those practices which are most directly required by JIT and TQM approaches. Inclusion will not be made in this study of other practices connected to human resource management such as recruiting, compensation, promotion and training, which are more influenced by human resource management policies, and which reflect the managerial vision of each individual company, the industrial context, the labour market, the cultural context, and the sector and country involved.
JIT is the organizational principle on the basis of which every working activity must be supplied with the necessary components in the necessary time and in the necessary quantity. In other words, in this way every element will reach the work station at the right time without having to be stored in intermediate warehouses. JIT therefore tries to give maximum value to the factory operations which actually generate product added value and to eliminate all forms of waste, starting from the functions which are not directly productive. Quality management is defined as an integrated approach to achieving and sustaining high quality output, focusing on the maintenance and continuous improvement of processes and defect prevention at all levels and in all functions of the organization, in order to meet or exceed customer expectations. Table I sets out some of the practices which characterize lean production as it has been described by various authors[1,4,16,22,36-40].
[TABULAR DATA OMITTED]
The work organization practices which are being taken into consideration are set out in Table II. Obviously, other aspects of the human resource management cycle such as selection, development and rewards are relevant since they influence the setting up and the maintenance of the work organization practices mentioned in Table II. The first step of the human resource management cycle is careful recruitment and selection of shopfloor workers, in order to develop a reliable and committed workforce. Potential employees should be screened for their task-oriented skills, their potential for working with a team, their dedication to quality values and their motivation and willingness to make improvements. Training of the workforce for small group problem solving, communication and statistical process control is necessary in order to develop the appropriate skills. The salary structure, the reward systems and the appraisal schemes certainly influence employee loyalty and commitment. Industrial relations, stability, and flexibility clauses in contracts which require trade union role that is closer to the Partner" than to the "adversary" are also essential for the vulnerable lean production system. In order to achieve flexibility important to have the ability within the company to develop and consolidate a nucleus of multi-skilled personnel (made up of intrinsically flexible suppliers of labour), and the ability to create numerically flexible external circuit of suppliers of labour, or, in other words, workers with a high level of skills standardization who can actually move between various companies[42-44].
[TABULAR DATA OMITTED]
In order to identify the work organization characteristics which most closely characterize lean production plants, the following hypotheses can be developed from the framework presented above.
H1:Lean production plants are characterized by greater employee commitment to continuous quality improvement.
Employee commitment to continuous quality improvement is the most well-known form of employee involvement. It is fundamental in order to keep processes under control and to improve them. It is therefore a prerequisite for both TQM and JIT.
H2:Lean production plants are characterized by greater use of small team problem solving.
The use of small team problem solving is one of the means available in order to improve the entire operative system. The solving of problems leads to quality improvement and greater control of production processes, and to their tighter coupling, thus obtaining smoother, faster and more reliable flows, all of these being typical aspects of a lean production system. The worker in a lean production plant also participates in the redefinition of the production process through his participation in problem solving.
H3: Lean production plants are characterized by greater consideration of worker suggestions for improvement.
As confirmed by various authors, it is to be expected that in lean plants the worker will contribute to the design/redesign of the production system through making appropriate suggestions. If the worker is to continue to make good suggestions and not repeat useless ones, it is essential him/her to perceive that his/her suggestions are taken seriously. It is important that he/she sees his/her suggestions being put into practice, or if not, that the reasons for the rejection of his/her proposal are explained to him/her.
H4: Lean production plants are characterized by greater supervisory interaction facilitation.
It is to be expected that the role of the supervisors is noticeably different in lean production. The supervisors are no longer simply controllers and schedulers of workers. In an environment where each member has to contribute to improving the system, the supervisors have the task of creating a climate which involves each individual and which encourages them to discuss their ideas.
H5: Lean production plants are characterized by greater contact between managers, engineers and workers.
It is to be expected that in lean production managers and engineers circulate on the shopfloor more often and talk to the workers more often in order to obtain direct information about production problems. This direct type of communication modifies the worker's perception of the significance of his/her own task. As a result of the increased presence of the managers and engineers, the workshop acquires greater importance for everyone and in addition to this there is a reduction in the workers' perception of hierarchical and functional barriers.
H6: Lean production plants are characterized by greater and faster feedback.
A lean production system has the characteristic of being able to adapt quickly to small variations in demand and of trying to reduce process variance. Greater and faster feedback directly to workers and supervisors are essential in order to achieve this systemic performance. The greater the quantity of feedback, the greater the number of aspects which can be kept under control, as well as the number of people who have the information necessary in order to take corrective action.
H7: Lean production plants are characterized by greater decentralization of authority.
Greater decentralization of authority makes it possible to handle uncertainty and improve the efficiency of the decision-making process. A system with more decentralization of authority should therefore be more reactive, a quality which seems appropriate to the lean production model. A greater decentralization of decision making is also inherent in teamwork and group problem solving. Moreover, participatory decision making has a positive impact on quality performance: one good example is the possibility of stopping the production flow due to quality problems[4,37].
H8: Lean production plants are characterized by greater use of multi-functional employees.
Workforce flexibility is required since JIT requires that idle time generated by rigidities in traditional job descriptions should be eliminated. JIT workers are therefore expected to be multi-skilled operators, who have the ability to run multiple machines, do their own quality control and solve quality problems[17,33,37]. The aim is to reduce work-in-progress and to minimize any idle equipment, facilities or indirect staff. In particular, it is to be expected that operators do their own quality control and document and improve their own performance, since this is associated with better quality performance[22,26,27]. It is also to be expected that workers should know how to maintain their own machines since lean production aims to adopt frugal technology (that is, to use simple machinery which can be accessible to the personnel using it) and at the same time it aims to reduce indirect personnel and increase each single operator's mastery of the process.
Workforce flexibility is particularly important so that the production system can be adapted to changes of mix and volume, and so that it is not sensitive to the absence of employees. Flexibility therefore aids production scheduling and work flow advancement also becomes smoother.
H9: Lean production plants are characterized by better documentation of shopfloor procedures.
In lean production the achievement of good process management and better documentation of shopfloor procedures is important in order to minimize the likelihood of operator error. Clear and up-to-date documentation also increases the flexibility of operators since they can more easily find out about and learn the actual activities to be carried out. Better documentation is also the result of small team problem solving, employees' suggestions and continuous improvement efforts. Documentation on production procedures is even a means through which the plant can acquire worthwhile knowledge and information. In order to test these hypotheses empirically one should have two groups of plants, one which makes considerable use of management systems typical of LP such as JIT and TQM and others which do not use them widely or indeed do not use them at all.
The empirical research approach followed is that of the survey. Very careful attention has been paid to sampling, approaching companies, and developing and testing measurement instruments since each one of these aspects can influence the results strongly This will increase the researcher's confidence in the meaning and the level of significance of the data used in testing the research propositions.
The population has been subdivided into two strata: companies which are well-known for their excellence (also known as lean producers) and the others on the basis of expert opinion. As regards the size of the company, the figure of 100 employees has been chosen as the lower limit. Particular attention was given to the contact strategy. Initial contact was made via formal letter to the company executive officer explaining the research objectives and assuring them feedback: 80 per cent of companies contacted agreed to participate, selected one of their plants and appointed a plant research co-ordinator, who functioned as the primary liaison officer for the project. Companies which did not accept were facing restructuring problems or belonged to the defence industry: the latter were worried about the problem of the confidentiality of data. Response was thus obtained over a total of 43 plants belonging to the electronics, auto supplier, machinery and mechanical sectors. Plants participating received a detailed profile of their performance and a sample profile for comparison.
Verification of the above research propositions through empirical research requires the measurement of involved variables. In Table III the characteristics of the measures used are presented and references to previous work which uses the same (or very similar) measures are reported in the first column. All the data necessary to assign a value to these variables were collected by means of structured questionnaires, even though plant visits were also carried out to deliver questionnaires and to get a general feeling about the plant through interviewing some managers. In developing measures for variables under study, both perceptual and objective questions were used. The answers to perceptual questions were measured on five points of the Likert scale. In the case of variables measured through perceptual questions, several questions for the same variable and several respondents for the same question from the same plant were used. The value for each plant was obtained by working out the average between the respondents on the same question and subsequently the average between questions relative to the same variable.
In total, 26 respondents per plant were involved from different functions and from different hierarchical levels. Twelve workers belonging to different transformation and assembly phases filled in three different questionnaires. Four supervisors filled in two different questionnaires. Then managers filled in different questionnaires. The managers involved were the human resource manager, the inventory manager, the purchasing manager, the information system manager, the production control manager, the process engineer manager, the plant manager, the plant superintendent, the quality manager and the plant research co-ordinator.
All the measures were then tested for their reliability and construct validity and items which undermined construct validity or construct reliability were eliminated[22,52]. Measure reliability is the property of the measure to be not affected by random error. The reliability analysis was done through the internal consistency method which is based on the mean correlation among the measure items. The coefficient Cronbach's alpha was used as the reliability indicator. Construct validity measures whether a set of items is an appropriate operational definition of an abstract variable or a construct. In order to have construct validity a set of items must represent one and only one concept, i.e. the common factor analysis on the set of items highlights only one significant factor. The number of significant factors is equal to the number of eigenvalues greater than the mean common variance between items. The percentage of common variance explained by the first factor can be used as an indicator of the validity level: high values indicate that the resulting factor collects a great part of the information shared between the couples of items. The percentage of total variance explained by the first factor is another useful indicator: high values indicate that the resulting factor can explain a large part of the information yielded by the set of items. For further details concerning the method see [47,54].
Table III reports the results of the reliability analysis. The second column shows the number of items used in the measure. The third column reports the number of respondents: the first number indicates the number of managers, the second one the number of supervisors and the third one the number of workers. The results of validity and reliability tests carried out at plant level (each item has only one value per plant) and at respondent level (the observation unit is the respondent, so each item has up to 13 values for one plant) are reported respectively in the fourth and fifth columns. In these columns the first number represents the Cronbach's alpha coefficient, the second one the percentage of common variance explained by the first factor, and the third one the percentage of total variance explained by the first factor. In the last two columns are reported respectively the mean and the standard deviation to provide the basic information on the distribution.
[TABULAR DATA OMITTED]
In order to test the hypothesis, a two sample unilateral t student test was carried out. In order to increase confidence in the results, even a distribution free test based on median scores was performed.
Before testing the hypothesis it could be useful to check if firms with an LP reputation are really more advanced in LP dimensions than firms with no LP reputation. Two tests were performed: a t-test and a median test. Tests were performed respectively using the TTEST and NPAR1WAY procedures of the SAS system. Generally the tests gave the same result. Table IV presents the results of the analysis. LP and NLP indicate respectively the mean value for the lean production and the non-lean production groups. A lower value indicates a lower presence of the practice. In the Table "prob. >F" indicates the probability of rejecting the null hypothesis of no difference between the variances of two groups when it is true; "prob. > T" indicates the probability of rejecting the null hypothesis of no difference between the means (the value reported considers if the group variances are equal or not) when it is true. As far as the median test is concerned "prob. > [absolute value of Z]" provides the information corresponding to "prob. > [absolute value of T]".
[TABULAR DATA OMITTED]
All the lean production practices considered actually distinguish LP reputation plants from other plants. This result gives greater reliability to the fact that the two groups of plants considered differ from each other as far as the lean production practices are concerned. The same analysis is performed on the work organization variables in order to test the hypotheses. Table V shows the results. H1 (even though at the limit), H3, H6 and H9 are definitely confirmed: Lean production plants seem to have employees that are more committed to continuous quality improvement, to take employees' suggestions more seriously, to rely more heavily on quality feedback to both workers and supervisors and to document production procedures very carefully.
The tests of H2 disagree. This is due to the presence of two outliers. On eliminating the two outliers it becomes clear that lean production plants use more teams for problem solving than the other plants.
H8 is partially confirmed since lean production plants show that their employees are able to perform a variety of tasks and in particular are able to perform statistical process control (SPC) themselves, but they do not show that they have more on-line workers able to perform maintenance. It should be noted that the two tests partially disagree as far as the workers ability to perform SPC is concerned. This is due to a high deviation from normal distribution (many plants have 0 per cent workers able to perform SPC). In any case, only LP plants have more than 30 per cent of workers able to perform SPC, and half of LP plants are actually in this situation. Even the H4 is partially confirmed: it does not seem that supervisors' role in favouring workers to act as a team is more active, even though suggestions are taken seriously. Finally, H5 and H7 hypotheses are not supported at all. Therefore, integration between management, engineers and workers and decentralization of authority (both in the general form and in the specific form of worker autonomation) does not seem to differentiate lean production plants from other plants.
This present study proposes a framework (developed from an analysis of the existing literature) which will be useful to research the linkages between work organization and lean production practices. The framework considers two types of work organization practices. The first one includes those practices which are closely linked to lean production practices such as JIT and TQM. The second deals with those practices such as recruiting, compensation, promotion and training which are strongly influenced by human resource management policies, and which reflect the managerial vision of each individual company, the industrial context, the labour market, the cultural context and the sector or country of origin. After this, the hypotheses concerning the practices which can be said to be closely linked to JIT/TQM are tested. By repeating the investigation in other contexts, it will be possible to obtain empirical identification of the work organization practices which cannot not accompany lean production practices, and this has immediate practical interest.
The result of this research is that lean production plants seem to use more teams for problem solving, to take employees' suggestions more seriously, to rely more heavily on quality feedback both for workers and supervisors, to document production procedures more carefully and to have employees able to perform a greater variety of tasks including statistical process control. Lean production plants show fewer differences, or even no differences, with regard to aspects of work organization which involve hierarchy: even though suggestions are taken more seriously it does not seem that the supervisors particularly favour workers to act as a team, and that integration between management, engineers and workers is not a differentiating characteristic, while decentralization of authority (both in the general form and with reference to the specific aspect of worker autonomation) is even less of a differentiating factor.
Some of the results obtained are unexpected. In particular some of the expected differences regarding work organization practices were not significant. Various conjectures can be made, which deserve further research in order to be supported or rejected. It can be assumed that the practices which do not turn out to be differentiated are not indispensable for a lean production system, or that some of the non-lean production plants are setting out along the road towards lean production by starting from work organization practices, or finally that some of the lean production plants are in a transitional phase as far as lean production practices are concerned. This last conjecture is supported by the fact that lean production initiatives were only recently introduced into some of the lean production plants considered. It could even seem that there is a priority order between the various work organization practices: some are indispensable for a lean system, while others adapt well to it but require more time and can be introduced later. Another explanation could be that the results are influenced by the country in which the plant is situated. In addition to these questions, there is a further big issue and this concerns the role of the human resource management practices which are needed in order to maintain the practices considered above in time.
[1.] Womack, J.P., Jones, D.T. and Roos, D., The Machine that Changed the World: The Triumph of Lean Production, Rawson Associates, New York, NY, 1990. [2.] Coriat B., Penser a l'Envers, Christian Bourgois Editeur, Paris, 1991. [3.] Bonazzi, G., Il tubo di cristallo. Modello Giapponese e Fabbrica Integrata alla Fiat Auto, Il Mulino, Bologna, 1993. [4.] Ohno, T., Toyota Production System, Productivity Press, Cambridge, MA, 1978. [5.] Bonazzi, G., "Modello Giapponese, produzione snella e miglioramento continuo", in Bonazzi, G. (Ed.), Storia del Pensiero Organizzativo, Franco Angeli, Milan, 1995, pp. 166-90. [6.] Revelli, M., "Introduzione", in Ohno, T. (Ed.), Lo spirito Toyota, Einaudi, Torino, 1993. [7.] Hall, J.L. and Leidecker, J.K., "Is Japanese-style anything new? A comparison of Japanese-style management with US participative models", Human Resource Management, No. 4, 1981. [8.] Oliver, N. and Wilkinson, B., The Japanisation of British Industry, Blackwell, Oxford, 1988. [9.] Litter, C.R. and Salaman, G., Class at Work: The Design, Allocation and Control of Jobs, Batsford, London, 1984, pp. 36-7. [10.] Huczinzinski, A. and Buchanan, D., Organizational Behaviour, 2nd ed., Prentice-Hall, London, 1991, p. 307. [11.] Wood, S., "Japanisation and/or Toyotism?", Work Employment and Society, Vol. 5 No. 4, December 1991, pp. 567-600. [12.] Groebner, D.F. and Merz, C.M., "The impact of implementing JIT on employees' job attitudes", International Journal of Operations & Production Management, Vol. 15 No. 1, 1994, pp. 26-37. [13.] Goyal, S.K. and Deshmukh, S.G., A critique of the literature on just-in-time manufacturing", International Journal of Operations & Production Management, Vol. 12 No. 1, 1992, pp. 18-28. [14.] Harber, D., Samson, D.A., Sohal, A.S. and Wirth, A., "Just-in-time: the issue of implementation", International Journal of Operations & Production Management, Vol. 10, No. 1, 1990, pp. 21-30. [15.] Hiltrop, J.M., "Just-in-time manufacturing: implications for the management of human resources", European Management Journal, Vol. 10 No. 1, 1992, pp. 49-54. [16.] Schonberger, R.J., Japanese Manufacturing Techniques: Nine Hidden Lessons in Simplicity, Free Press, New York, NY, 1982. [17.] Klein, J., "The human cost of manufacturing reform", Harvard Business Review, March-April 1989, pp. 60-66. [18.] Wilkinson, B. and Oliver, N., "Human resource management in Japanese manufacturing companies in the UK and USA", in Towers, B. (Ed.), The Handbook of Human Resources Management, Blackwell, Oxford, 1992. [19.] Bratton, J. and Gold, J., Human Resource Management. Theory and Practice, Macmillan, London, 1994, pp. 57-83. [20.] Burawoy, M., Manufacturing Consent: Changes in the Labour Process under Monopoly Capitalism, Chicago University Press, Chicago, IL, 1979. 21. Heard, J.A., "Management policy changes: a must for JIT", 1986 APICS International Conference Proceedings, 1986, pp. 304-09. [22.] Flynn, B.B., Schroeder, R.G. and Sakakibara, S., "A framework for quality management research and associated measurement instrument", Journal of Operations Management, Vol. 11 No. 4,1994, pp. 339-66. [23.] Wilkinson, B. and Oliver, N., "Power, control and the kanban", Journal of Management Studies, Vol. 26 No. 1, 1989, pp. 47-58. [24.] Ashford, S.J. and Cumming, L.L., "Feedback as an individual resource: personal strategies of creating information", Organizational Behaviour and Human Performance, Vol. 32, 1984, pp. 370-98. [25.] Ilgen, D.R., Fisher, C.D. and Taylor, M.S., "Consequences of individual feedback on behaviour in organizations", Journal of Applied Psychology, Vol. 64, 1979, pp. 349-71. [26.] Greller, M.M. and Herold, D.M., "Sources of feedback: a preliminary investigation", Organizational Behaviour and Human Performance, Vol. 13,1975, pp. 244-56. [27.] Hanser, L.M. and Muchinsky, P.M., Work as an information environment", Organizational Behaviour and Human Performance, Vol. 21,1978, pp. 47-60. [28.] Juran, J. M. and Gryna, F.M., Quality Planning and Analysis, McGraw-Hill, New York, NY, 1980. [29.] Aoki, M., Information, Incentives, and Bargaining in the Japanese Economy, Cambridge University Press, New York, NY, 1991. [30.] Fucini, J.J. and Fucini, S., Working for the Japanese, Free Press, New York, NY, 1990. [31.] Wilkins, A.L. and Ouchi, W.G., "Efficient cultures: exploring the relationship between culture and organizational performance", Administrative Science Quarterly, Vol. 28, 1983, pp. 468-81. [32.] Dutton, B., "A case for work teams", Manufacturing Systems, Vol. 9 No. 7,1991, p. 58. [33.] Aggarwal, S.C., "MRP, JIT, OPT, PMS?" Harvard Business Review, September-October 1985, pp. 8-16. [34.] Hackman, J.R. and Oldham, G., Work Redesign, Addison-Wesley, Reading, MA, 1980. [35.] Kelly, J., "Does job re-design theory explain job re-design outcomes?", Human Relations, Vol. 45 No. 8,1992, pp. 753-74. [36.] Schonberger, R.J., World Class Manufacturing, Free Press, New York, NY, 1986. [37.] Monden, Y., Toyota Production System: Practical Approach to Production Management, Institute Engineering and Management Press, Institute of Industrial Engineers, Norcross, GA, 1983. [38.] Gunn, T.G., Manufacturing for Competitive Advantage: Becoming a World Class Manufacturer, Ballinger, Cambridge, MA, 1987. [39.] Schonberger, R.J., Building a Chain of Customers: Linking Business Functions to Create the World Class Company, Free Press, New York, NY, 1990. [40.] Giffi, C., Roth, A.V. and Seal, G.M., Competing in World Class Manufacturing, Business One Irwin, Homewood, IL, 1990. [41.] Weiss, A., "Simple truths of Japanese manufacturing", Harvard Business Review, Vol. 62, 1984, pp. 119-25. [42.] Dei[beta], M., "Toward a polarisation of skill structures", in Altmann, N., Kohler, C. and Meil, P. (Eds), Technology and Work in German Industry, Routledge, London, 1992. [43.] Kumazawa, M. and Yamada, J., "Jobs and skills under the lifelong Nenko employment practice", in Wood, S. (Ed.), The Transformation of Work? Unwin Hyman, London, 1989. [44.] Cainarca, G.C., "Gli uomini, le macchine e l'organizzazione del lavoro", in Mariotti, S. (Ed.), Verso una Nuova Organizzazione della Produzione. Le Frontiere del Post-Fordismo, ETASLIBRI, Milan, 1994, pp. 99-140. [45.] Elmuti, S.D., Quality control circles in Saudi Arabia: a case study", Production and Inventory Management Journal, Vol. 30 No. 4,1988, pp. 52-5. [46.] Nohara, H., "Technologies Electroniques et Gestion de la Main-d'oeuvre dans l'Industrie Japonaise", Revue d'Economie Industrielle, Vol. 34 No. 4, 1985. [47.] Forza, C., "Quality information systems and quality management: a reference model and associated measures for empirical research", Industrial Management & Data Systems, Vol. 95 No. 2, 1995, pp. 6-14. [48.] Georgopoulos, B.S. and Mann, F.C., The Community General Hospital, Macmillan, New York, NY, 1962. [49.] Misterek, S.D.A., Bates, W.A., Morris, WT. and Schroeder, R.G., "Manufacturing strategy and organizational culture: theory and management issues", in Voss, C. (Ed.), Manufacturing Strategy - Theory and Practice, Proceedings of the 5th International Conference on the OMA UK, MCB, Bradford, 1990. [50.] Taylor, J.C. and Bowers, D.G., Survey of Organizations: A Machine Scored Standardized Questionnaire Instrument, Institute for Social Research, University of Michigan, 1972. [51.] Hage, J. and Aiken, M., "Relationships of centralization to other structural properties", Administrative Science Quarterly, Vol. 12, 1967, pp. 72-92. [52.] Saraph, J.V., Benson, P.J. and Schroeder, R.G., "An instrument for measuring the critical factors of quality management", Decision Science, Vol. 20 No. 4, 1989, pp. 810-29. [53.] Cronbach, L.J., "Coefficient alpha and the internal consistency of tests", Psychometrika, Vol. 16, September 1951, pp. 297-334. [54.] Flynn, B.B., Sakakibara, S., Schroeder, R.G., Bates, K.A. and Flynn, E.J., "Empirical research methods in operations management", Journal of Operations Management, Vol. 9 No. 2, 1990, pp. 250-84. [55.] SAS Institute, "SAA/STAT Volume 2. Version 6 Fourth Edition", SAS Institute, Cary, NC, 1990, pp. 1195-210; 1633-40.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Lean Production and Work Organization|
|Publication:||International Journal of Operations & Production Management|
|Date:||Feb 1, 1996|
|Previous Article:||Assessing changes towards lean production.|
|Next Article:||New wave manufacturing strategies: operational, organizational and human dimensions.|