Transplanting Japanese success in the UK.

Producing inside the European Community offers various advantages to Japanese firms, many of them related to costs. In manufacturing, for example, it enables them to avoid paying tariffs, to avoid quantitative restrictions, to reduce shipping costs and to obtain better market access.(1) To make

the exploitation of those cost advantages worthwhile the production cost differences between a Japanese company's plant in Japan and a plant in Europe should not exceed the advantages to be reaped. It is thus reasonable to expect that while a Japanese transplant in Europe may not be able to achieve parity with cost and productivity levels in its parent it can nevertheless outperform many UK plants in the same industry if the parent was a competitive exporter.(2) Some of the better performance will come from purely technical factors such as more advanced and higher quality products, more efficient production processes and access to higher quality components, all of which can be brought in from outside Europe. However, even after taking account of those 'technical' factors efficiency differences remain, largely related to the quality and utilisation of the labourforce.(3)

In the research project discussed in this note we investigated the extent to which Japanese transplants were able to achieve improved productivity and lower costs in the UK by transplanting Japanese management methods which entailed changing how UK workers and firms behaved. We examined various of the aspects of Japanese labour force management which are widely believed to improve productivity performance, such as training, teamworking, lifetime employment and bonus payments.(4)(5)

The results are based on a series of interviews held with Japanese and UK managers in 10 transplants, primarily in consumer electronics, other electronic products and vehicle manufacturing. We also conducted some interviews with service companies, primarily financial institutions and hotels to gain some insight into the extent to which this transfer of management practices was largely dependant upon the specific nature of the manufacturing process. Although the sample is small it represented about a quarter of the relevant population. Selection of the sample was non-random, being chosen deliberately to give a wide product and geographical coverage. Seven of the companies were greenfield investments, while the remainder were acquisitions. The plants had been in production for between 3 and 15 years so that we could form a judgement about the rate at which the learning process could take place.

The plants were selected from the DTI and JETRO lists of Japanese investments in the UK.(6) The UK plants were approached directly (in Japanese), not through the head offices, and a 100 per cent response rate was achieved. The investigation was short. After collecting such published data as were available the plants were sent a brief questionnaire to prepare for a visit on the detail of the products, production processes and workforce. We then interviewed senior Japanese and British managers (each in their own language) using a structured interview schedule. With one exception, the interviews took place during October, 1991 to February, 1992, each lasted half a day and involved a minimum two and a maximum of five managers.

Productivity levels in the transplants and their parents

Each of the companies we visited were using Japanese machinery of the same type available in the Japanese plants. However, in some cases it was the previous vintage and the current plant in Japan had more modern equipment.(7) In others, although the equipment was identical, it was not necessarily run at the same speed (in one case a particular assembly process was run at half the Japanese rate). In all but three cases productivity was lower in the UK than in Japan and in only one were there examples of higher UK productivity. However, there was also only one case where the UK productivity levels more than 15 per cent lower. Since this plant had been running for around 10 years, the differences could not be ascribed merely to being in an early stage in the learning process. Quality levels were generally lower than in Japan and nowhere higher but the discrepancies were much smaller than they were for productivity.

To an extent the degree of similarity is exaggerated by these remarks as the range of products produced was lower in the UK and the extent of automation somewhat lower. To a degree this choice of a lower capital labour ratio reflected the lower relative cost of labour in the UK, not just differences in labour efficiency which required higher manning levels. The Japanese managers interviewed suggested that there was no obvious difference in performance between UK and Japanese workers who had acquired similar levels of skills and experience. However, in general, skill and experience levels were lower in the UK both in terms of the external skills acquired before employment in the plant but also in terms of learning on the job within the plant. No attempt had been made to compare rates of learning by nationality although some employees had been sent on common training programmes in Japan with the employees of transplants in other countries.

Differences in skill levels were cited by most of the plants as part of the explanation of differences in productivity but the sources of non-human differences: differences in machines, quality of components, production mix and computer control systems, related only to two or three plants in each case, so there were no strong common themes. The companies had, however, experienced a number of difficulties with component suppliers. Many crucial components, such as integrated circuits, transistors, diodes and resistors, had been imported from Japan, even where local content was as high as 70 per cent, as this helped maintain input quality and security of supply. The large majority of local (i.e. European) content came from the UK (95 per cent in one instance) and although there had been very considerable success in getting UK suppliers to meet the purchasers' requirements, the problems had been eased by the arrival of some of the Japanese suppliers with their own transplants in Europe. The preference for relying on traditional links was very strong; in the case of one financial intermediary, it extended to employing a Japanese removal firm when it moved between buildings in London.

Differences in the workforces

In an attempt to assess how much the workforce behaviour in the transplants fell short of that in the parents and how much it exceeded that of other plants in similar industries in the UK, we asked a short set of questions of all the interviewees. Following the work of Prais (1990) and earlier work on the sources of Japanese productivity differentials (cited above), we concentrated on two main areas: the acquisition and exercise of skills and the motivation of employees to work together and in the interests of the company. The twelve main questions are set out in Table 1. These questions are deliberately relative. The Japanese managers were asked to rate their current UK employees against their experience from plants in the same company in Japan, while the British managers were asked to rate their current employees relative to those they had managed in similar companies elsewhere in the UK. (Managers who had not previously worked in a similar Japanese plant/the same industry in the UK were excluded.)

A very simple 5 point rating scale was used, as set out in the footnote to the table, and results were averaged with uniform weights, first within the plant and then across the different plants to produce the figures in the table. All the usual caveats about lack of comparability of subjective views apply. In addition there are some obvious sources of bias. The comparisons are all with previous, not contemporaneous, experience. Even the most honest manager is unlikely to want to describe his current workforce too unfavourably as that might imply that his (all the interviewees were men) management was in some sense ineffective. This may help explain the lack of extreme values.

Table 1. Assessments of workforce characteristics

Characteristic                                   Rating scales
                                                 (J)       (L)

1  speed and accuracy of work                    2.2       3.7
2  speed of learning and skill acquisition       2.9       3.5
3  skills in operating machinery                 2.7       3.6
4  problem solving ability                       2.0       3.3
5  knowledge of products, processes &
machinery                                        2.5       3.7
6. willingness to improve working environ-
ment                                             2.2       3.7
7  co-operation and flexibility in teamwork      2.4       4.3
8  competitive spirit                            2.8       3.6
9  willingness to improve skills & knowl-
edge                                             2.8       3.7
10 sense of belonging & participation in
company                                          1.7       3.9
11 diligence, willingness to work                2.2       3.5
12 sense of responsibility for job & product     2.2       3.8
(unweighted average                              2.4       3.7)

Rating scale: 5=much better; 4=better; 3=the same; 2=worse;
1=much worse.

Nevertheless the table has some interesting characteristics. First of all, average rankings by the Japanese managers (J) are uniformly lower than of the equivalent plant in Japan, while the rankings by the British managers (L) are uniformly higher than in equivalent plants elsewhere in the UK. However, it is interesting that learning ability (line 2) and the drive both to acquire skills and perform better than others (lines 9 and 8) were very little different from Japan. Similarly, these characteristics were not especially highly rated by the UK managers compared with their experience from other companies. This implies that British workers have the potential to acquire the same skills as their counterparts in Japan (although of course the starting points, and hence the costs, will be different).

At the other end of the scale the ability to solve problems, typically to repair faults or institute preventative maintenance, (line 4) was lowly rated as was the ability to identify with the company (line 10). This first finding conforms strongly with the work of Prais et al. as summarised in Prais (1990). However, it is noticeable that the UK managers also rank this lowest, implying that the transplant has been relatively unsuccessful in changing this attribute. However, the company is not merely taking on innate abilities, it is taking on workers, trained and educated in their respective systems. The Japanese workers will typically have worked in the same company for some time, whereas the UK workers will necessarily have obtained their experience elsewhere. They may therefore have bad habits to unlearn as well as a very different level of basic skills. (Work on skill and qualification levels is being undertaken in a separate and much larger study by Mari Sako at the LSE, so we avoided duplicating it.)

However, the sense of identification with the company is highly rated by UK managers compared with their experience of other plants. Thus this characteristic, which is thought to be an important aspect of Japanese success, shows that an element of transfer has taken place but there is a lot more that can be achieved before parity is reached with Japanese plants. The same can be said of co-operation and teamwork (line 7) and the sense of responsibility that the employee exhibits for his own actions and the products of the company (line 12). These are highly rated compared to other UK companies, suggesting that these messages have successfully been put across but still rated well below experience in Japan. Thus there has been some successful transfer of the Japanese working culture. The emphasis on the team is a most important difference. Respondents argued, for example, that in a traditional British firm, an employee discovering how to perform a task better would tend to keep that information to himself or otherwise use it so as to boost his personal standing with the company relative to his fellows. Whereas in the 'Japanese' environment, the expected behaviour would be to share the knowledge with everyone else in the team, so that the company could benefit as soon and as widely as possible, rather than the individual. In any case it would be expected that the origin of such ideas would be known and rewarded in due course.

Inevitably the rankings will reflect not just relative judgements but also views of the absolute importance of particular characteristics. Where a particular factor receives emphasis within the firm it is likely to be rated as an area of Japanese advantage by the Japanese managers and one of transplant advantage by their British counterparts.

The interviews explored several aspects of these differences in labourforce characteristics in greater detail but in this note we focus on just four of them: skills, training, incentives and appraisal; as these attracted most comment by the Japanese managers in their assessment of differences between the UK and Japanese workforces.

Skills

The gap in skills and knowledge between the two workforces was cited(8) as the single most important factor differentiating the two workforces. The UK workers are not of course a cross-section of the UK workforce at large either in the specific industries or in the economy as a whole. The greenfield transplants had the advantage of being able to choose among a wide pool of skilled labour typically during a period of high unemployment and job insecurity. They were therefore able to pick from among the better workers in the industry rather than merely recruit those whom the existing employers did not want. Nevertheless, at the shopfloor level most of the employees were not recruited from UK competitors but from firms in other industries or from the unemployed. In brownfield investments the new Japanese employer has to cope with retraining and adapting an existing labourforce, which presents different problems. However, although the aspiring acquirer will find it difficult to takeover the very best companies in the industry, it will not be interested in the worst but concerned to find good companies which are underperforming and capable of improvement. Again, therefore, the workforce may not be representative of the full range of UK experience, although it could be 'typical' for the industry.

Even if basic skills were identical, firm-specific skills would have to be acquired by the UK labourforce, which would typically take years rather than months. As labour turnover in the transplants was in general low, down to Japanese levels in a couple of cases, most of the plants had had the opportunity to expose their workforces to a sufficient length of training that they could be expected to reach a considerable proportion of the parent's skill levels. Of course, some of the Japanese employees nearing retirement will have been with the company much longer, so the experience cannot be fully comparable. Our study distinguished between four general groups of employees: operatives, supervisors, engineers and technicians, and managers. The skill gap and the ability to handle it varied considerably among these groups.

The operative group was largely defined in terms of relatively repetitive tasks, for which the range of required skills was limited and the learning curve short and steep. Here the Japanese managers found that there was little problem in finding workers with an adequate skill levels or in training them up rapidly to the required standard. The problem came in the more skilled categories. Supervisors provided the worst difficulty. A good supervisor is thoroughly familiar with the problems of the production line, having been recruited from it and working closely with it subsequently. A supervisor is an intermediary between management and shopfloor, requiring some management skills and yet being able to identify with the problems of the workers being supervised and indeed represent their interests to management (in part therefore taking on what might be regarded as a trade union or shop steward role in a traditional UK plant).

This involves having both engineering and administrative skills. A good supervisor needs to be familiar with the technology, the specific machines, the production practices and be good at human relations as well to provide the appropriate motivation and teamworking spirit. These combinations were almost impossible to find. In general, potential supervisors had a lower level of general education than in Japan, which meant that the range of tasks they could cover was smaller. In so far as they had technical and engineering knowledge it tended to be specialised and this was the way that they had managed to transcend from the group of more general workers in their previous employment. Thirdly those who were supervisors expected to move away from the operative tasks and not to be involved with repair and maintenance themselves but to have an administrative style job.

Although we have lumped engineers and technicians together there was an important practical distinction in the companies we looked at, in that they expected to be able buy in engineers who already had adequate skills, while technicians could be largely trained internally, on the job. This therefore provided a problem in the case of engineers as there had appeared to be a shortage. Engineers frequently had to be lured away from competitors and had mobile skills, so that exceptional measures had to be taken to encourage them to remain with the company, thus distorting the system away from the considerable equality of rewards for administrative, production and engineering staff that existed in Japanese plants.

Interestingly enough management skills provided a rather different problem. A typical Japanese manager would be rotated through a wide range of jobs within the company on his way up the ladder. Not only did new British employees not have that experience within the same company as their Japanese counterparts but the scope for such rotation did not exist. Such rotation would require excellent command of the Japanese language and the willingness to work in Japan as the UK plants were smaller and had a more limited range of functions. Furthermore Japanese employment legislation would make such actions difficult even if the employee were suitably qualified and willing. Some scope existed in some companies for rotation with North America, however, in general there was no particular encouragement to learn Japanese or to integrate managers into the full Japanese system.

This meant that there was inevitably a clear qualitative difference between the transplants and their parents. Non-Japanese managers could not participate fully in the development of decisions within the firm nor be a full party to the network of linkages which characterises many companies. In part this enforces a rather larger degree of autonomy on the UK subsidiary, especially in the case of service industries, unless very extensive use is to be made of Japanese expatriates, which was not normally the strategy.

Training

Given the gap between skill requirements and skills available in the UK workforce, training had a high priority. The transplants were faced with rather different training needs than their Japanese parents. First of all the employees came from a very different cultural background and hence could be expected to react in a different way to many common incentives. In part a training programme has either to overcome those cultural differences or to adapt the Japanese way of working to allow for them. In practice the firms chose a middle course, adapting in both respects.

Training can take place within or outside the company but the bulk of acquisition of skills we were told about rested on the internal route (kigyo-nai kyoiku). Extensive use was made of on the job training, with experienced workers assisting those who were new to the job. At its best this learning by example approach fed smoothly into teamworking where each expected to learn from his peers and pooling of knowledge and information took place on an almost continuous basis, despite rather than because of specific opportunities set aside for the discussion of improvements that could be made. External training was required mainly for various general skills which might have been expected of an equivalent Japanese worker, particularly those necessary to take both technical and managerial decisions. Despite the considerable training undertaken, it proved largely impossible to cover all the gaps. More senior staff were sometimes taken to Japan for a more thorough introduction to the company philosophy and ways of operating. Seeing the ideas actually put into practice in Japan was thought to be a very effective contributor to the process of acquisition for all levels of employee.

The rate of knowledge acquisition by UK workers was broadly comparable with that in Japan. It was surprisingly easy to break some existing habits and increase flexibility but several of the Japanese managers remarked that there was still a greater tendency to seek excuses for taking a break from work, whether during a breakdown or a shortage of components than was common in Japan. For those jobs where the amount of knowledge required was relatively limited, knowledge and skill levels could be brought up to Japanese levels but cultural differences in application, identification with the company and the product, etc. tended to mean that some gaps still remained. One interesting difference was the desire by the UK employees to obtain externally recognised qualifications, thought by the Japanese management to reflect a wish to have transferable skills, which would enable them to move to other companies in the industry should the need or desire arise.

Incentives

The Japanese system in the larger companies offers considerable incentives to the employee from a number of points of view. First of all there is the expectation of lifetime employment. Secondly a large proportion of salary is paid in the form of bonuses dependent upon the performance of the whole unit, not just the individual within it. Hence there is an incentive for all to encourage the weakest members in the team as all would suffer financially otherwise. However, the largest part of the bonus relates to the company as a whole. While this may increase loyalty to the company, it reduces the extent to which the individual employees can affect their own remuneration.

The general grade structures applied in Japan existed in the plants we visited, with five main grades (Senior manager, bucho; Manager, kacho; Supervisor, kakaricho; Team leader, hancho and Operator, ippan or hira) and subcategories within them. Pay structures, however, tended to be rather simpler in the transplants, with salary depending largely upon grade, which is fixed on an annual basis after an appraisal of performance. Elements of salary relating to the particular job, group and individual bonuses and length of service are less common. Hence much of the Japanese incentive structure is missing. In part this appears to be the result of a worry that it would result in too big a discrepancy between the pay structures of the transplants and other UK producers. There was thus little evidence available either of the effectiveness of the various incentive schemes or of the degree to which such schemes would transferable to the UK context. This was rather disappointing as evidence from Davies and Lyons at the University of East Anglia suggests that, unlike most multinationals, Japanese multinationals tend to pay a little less than the UK average.(9) This should imply that other characteristics of the system should stand in lieu of that lower pay, e.g. non-pecuniary advantages or greater job security. However, in practice it is rather difficult for that security to be offered, as most companies have only one transplant in the UK and shifting of employees between foreign countries is far less like security than shifts between plants or jobs in Japan.

The combination of influences meant that junior employees in the UK were relatively better paid compared with the average in the plant than their Japanese counterparts. However, at the other end of the scale, salaries did not rise so steeply with seniority, giving a somewhat flatter profile in the UK--in part perhaps explained by the more limited long run opportunities.

Appraisal

This system is completed by the annual appraisal, where each employee is rated by his direct superior according to a set of criteria, which include not just efficiency in performing the direct tasks which face them but attendance, attitudes, relations with others, knowledge and skills, etc. On this basis the grade is fixed for the next year, which is the major component of pay. However, it is not a one-sided arrangement as the superior has to agree the appraisal with the person reported on. In Japan self appraisal is common but only one plant in the UK followed this route. Much of the point of this exercise is to determine the needs for career development although it was the exception rather than the rule in the transplants.

Discussion

Our small enquiry did indeed suggest that it has been possible to transplant more than just the 'technical' factors which contribute to higher Japanese productivity, such as the quality of components, sophistication of machines, improved control systems, etc. Japanese systems of training, teamworking and operational flexibility have been introduced. In some plants it has been possible to reach Japanese levels of productivity but the general experience has been that, although it has been possible to achieve considerable advances over the performance of equivalent UK companies, there is still some ground to be made up before the Japanese levels are reached.

Some aspects of the cultural background are impossible to transfer. Employees will come from a UK not a Japanese educational system, their peers will still be working for non-Japanese firms. As time goes past the problems of having to take on staff with experience in other environments will diminish, although the problems of small size compared with the Japanese operations will always increase the dependence on outside trained staff. Parts of the incentive and experience systems cannot be incorporated. Employees cannot be shifted through a range of jobs requiring fluency in Japanese nor would lifetime employment guarantees be feasible. However, other parts of the system, with group bonuses, a greater emphasis on seniority and teamworking would be possible. As the success of the culture becomes more obvious and is extended to suppliers in order that quality and delivery targets can be met, so will the sources of this advantage be emulated.(10)

While emulation may reduce the relative advantages that the transplants enjoy this will still be to the benefit of the consumer and the economy as a whole as efficiency will be increased. In this way the transplants make a direct contribution to the UK economy by increasing output and employment (most of the current transplants have displaced very little pre-existing UK capacity and have not been on sufficient a scale to have a macroeconomic effect - although this will change in the case of motor cars as the UK becomes a net exporter again). However, they also make an indirect contribution through increasing the skills of the workforce both internally and in their suppliers. This stimulates others into action as well as providing a base from which staff could move to other companies (despite the disincentives in the Japanese system to intercompany movement). The second part of our study, not reported on here, examines the success of the transplants in transfering their methods to other companies through the supply chain (not just to suppliers but also to distributors and service agents).

Transplants relating to the final stages of the production process are also tending to encourage Japanese component suppliers to set up in the UK, thus spreading the sphere of influence. Our study relates primarily to manufacturing but these changes are also occurring in services, in some cases in industries like financial services, where it was widely felt that it was the UK which had the better performance, not Japan.(11) In some respects it is these Japanese sources of influence which have been more effective than the opportunities and threats offered by the Single European Market in stimulating change in the UK economy but that Market has itself contributed to the incentive for Japanese firms to transplant here. As the Single Market develops the incentive will continue. If the early transplants are successful and the culture spreads this may result in process of cumulative causation, continuing to attract a disproportionate share of this investment to the UK.(12)

REFERENCES

Aoki, M. (1984), The Economic Analysis of the Japanese Firm, Amsterdam: North Holland.

Balasubramanyam, V.N. and Greenaway, D. (1992), 'Economic integration and foreign direct investment: Japanese investment in the EC', Journal of Common Market Studies, vol. 30, no. 2, June.

Barrell, R.J., Pain, N. and Lozachmeur, V. (1992), 'A panel data analysis of Japanese direct investment in Europe', mimeo, National Institute of Economic and Social Research.

Button, K. and Mayes, D.G. (forthcoming), 'Technical Inefficiency in Manufacturing Industry Revisted' mimeo, University of Loughborough and National Institute of Economic and Social Research.

Clark, R. (1979), The Japanese Company, New Haven: Yale University Press.

Cole, R. (1979), Work, Mobility and Participation, Berkeley: University of California Press.

Dore, R. (1973), British Factory--Japanese Factory, Berkeley: University of California Press.

Drake, T.A. and Caves, R.E. (1992), 'Changing determinants of Japanese foreign investment in the United States', Journal of the Japanese and International Economies, vol. 6.

Dunning, J.H. (1973), 'The determinants of international production', Oxford Economic Papers, vol. 25, pp. 289-336.

Dunning, J.H. (1986), Japanese Participation in British Industry, Beckenham: Croom-Helm.

Export-Import Bank of Japan (1992), Results of FY1991 Foreign Direct Investment Survey, Research Institute of Overseas Investment, London.

Gittleman, M. and Dunning, J.H. (1991), 'Japanese multinationals in Europe and the United States: some comparisons and contrasts', in (M W Klein and P J J Welfens, eds.), Multinational Enterprises in the New Europe and Global Trade, New York: Springer.

Hart, P.E. (1992), 'The Effects of 1992 on the Insurance Industry in Britain and Germany', National Institute Discussion Paper, no 18.

Hashimoto, M. (1979), 'Bonus payments, on-the-job training and lifetime employment in Japan', Journal of Political Economy, vol. 87, October, pp. 1084-1104.

Horst, T. (1971), 'The theory of the multinational form: optimal behaviour under different tariff and tax rules', Journal of Political Economy, vol. 79, pp. 1059-1072.

JETRO (1984), 'Productivity and quality control: case studies' in Sato and Hoshino (1984), pp. 197-215.

JETRO (1991), Seventh Survey of European Operations of Japanese Companies in the Manufacturing Sector, JETRO, London.

Japanese Management Association (1989), Kanban: Just-in-Time at Toyota (D Lu transl.), Cambridge, Ma: Productivity Press.

Kogut, B. and Chang, S.J. (1991), 'Technological capabilities and Japanese foreign direct investment in the United States', Review of Economics and Statistics, vol. 75, pp. 401-413.

Koike, K. (1984), 'Skill formation systems in the US and Japan: a comparative study', in Aoki (1984), pp. 47-75.

Langhammer, R.J. (1991), 'Towards Regional Entities in Asia-Pacific: the role of Japanese foreign investment in service industries', ASEAN Economic Bulletin, vol. 7, no. 3, (March), pp. 277-289.

Mayes, D. and Green, A (1992), 'Technical inefficiency in U K manufacturing industry', ch. 4 in R E Caves Industrial Efficiency in Six Nations, Cambridge Ma: MIT Press.

Mincer, J. and Higuchi, M. (1988), 'Wage structures and labour turnover in the US and Japan', Journal of the Japanese and International Economies, vol. 2, pp. 97-133.

Norman, G. (1992), 'Japanese Foreign Direct Investment: the impact on the European region', mimeo, University of Leicester.

Prais, S.J. (1990), Productivity, Education and Training: Britain and other countries compared, National Institute of Economic and Social Research.

Sato, K. and Hoshino, Y. eds. (1984), The Anatomy of Japanese Business, New York: M E Sharp.

Taiichi, O. (1984), 'How the Toyota production system was created' in Sato and Hoshino (1984), pp. 385-403.

Thomsen, S. (1992), 'Japanese direct investment in the European Community: the product cycle revisited', Chatham House, mimeo.

Thomsen, S. and Nicolaides, P. (1991), The Evolution of Japanese Direct Investment in Europe: death of a transistor salesman, London: Harvester-Wheatsheaf.

Vernon, R. (1966), 'International investment and international trade in the product cycle', Quarterly Journal of Economics, vol. 80, May.

NOTES

(1) There is a considerable literature on the determinants of direct investment by Japanese companies at the aggregate (Barrell et al., 1992; Thomsen, 1992; Balasubramanyam and Greenaway, 1991), industry (Drake and Caves, 1992; Kogut and Chang, 1991) and firm (Dunning, 1973, 1986; Gittleman and Dunning, 1991; Horst, 1971) levels. The motivations identified are several (JETRO, 1991; Export-Import Bank of Japan, 1992). Only some relate directly to cost. Others relate to the need to have control to ensure co-ordination and quality, or to the wish to respond to the moves of competitors in highly oligopolistic industries (Norman, 1992). Direct investment is often a complement to exporting, making it easier, and not just a substitute. (It may, for instance, help complete the product range.) However, this note focusses on issues relating to costs and efficiency rather than strategic market behaviour, although the latter cannot be ignored in explaining the persistence of differences. Questions relating to the choice of location are less relevant, however, as, in this analysis, the choice has already been made and the discussion relates to some of the consequences.

(2) Insofar as the main motivation for local production is market access then considerable cost differences could be maintained subject to the expectation of a satisfactory pay-off in the long run.

(3) See Mayes and Green (1992) and Button and Mayes (forthcoming).

(4) See, for example, Aoki (1984); Clark (1979); Cole (1979); Dore (1973); Hashimoto (1979); Koike (1984); Mincer and Higuchi (1988).

(5) Further parts of the study, to be completed, look at the relationship between the firm and its suppliers and the introduction of Just-in Time supply, Total Quality Management and similar means of reducing the costs of working capital and wastage. The way these gains can be exploited is illustrated in JETRO (1984), JMA (1989) and Taiichi (1984).

(6) These mimeographed documents are updated periodically and were intended by their compilers to be complete coverages up to within a year of our use of them.

(7) Some companies appeared to be practising a product life cycle approach to the location of production (Thomsen, 1992; Vernon, 1966) developing the product and producing it initially in Japan and then progressively locating in plants in other markets.

(8) The remarks in this and the following three sections refer to the findings from the company interviews.

(9) Unpublished data provided to the National Economic Development Office, 1988.

(10) As Langhammer (1991) has noted for parts of East Asia, this culture can become widespread and attract other Japanese investors because the learning costs in their plants will fall, with employees and suppliers starting from a higher base than was the case when the first investments took place.

(11) However, there is a lack of empirical evidence to substantiate this claim and the sheer size of Japanese financial intermediaries (see Hart (1992) in the case of insurance) must leave this open to question.

(12) As argued by Norman (1992). Norman goes on to show that although Japanese companies have stated in intentions surveys that a larger proportion of investment will go Germany, in practice this has not occurred. He suggested that the intention may reflect the expectation that since Germany is at the core of the Community market and has a high skill base it will be a desirable location. Whereas the subsequent action reflects a wider range of considerations which are much more favourable to the UK (JETRO, 1991).