model of the typical Japanese firm that contrasted with the typical model of an American firm. Aoki’s aim was to connect studies from diverse disciplines on the industrial organisation of the Japanese firms into a consistent and unified economic theory.
This paper will begin with a detailed explanation of Aoki’s model of the J-firm. A quite detailed version is necessary to highlight some of the misconceptions that have arisen as people have blurred Aoki’s actual theory with prior stereotypical conceptions of the Japanese firm. Then Aoki’s theory will be examined and evaluated. The core of examination will be on the J-firm model rather than Aoki’s conceptualisation of the A-firm.
The difference between the A-firm and J-firm are predominately in the three aspects: industrial organisation, labour relations and corporate financial structure.
Within the A-firms workers conduct specific tasks as per a job classification scheme postulated in a collective agreement. When irregular events happen, such as machine malfunction, remedies are sought by supervisors, engineers or other specialists. Efficiency is to be attained through job specialisation and rational hierarchical control.
In the J-firm workers’ roles are not specified in detail and the is frequent job-rotation. As a result they are familiar with the whole process and have the ability to cope with unexpected events. This signifies a delegation of decision-making power to shop floor workers. Workers’ are able to cope with irregular events through collective learning and knowledge leveraging. To Aoki such a system is effective in the automobile and steel industries, but with high-tech manufacturing he was unsure.
Decisions in the A-firm are hierarchically organised. Decisions that involved a high degree of uncertainty, for example in investment or R&D, are also under hierarchical control. However once the overall production framework has been imposed, horizontal informational exchanges and co-ordination of operations by relevant subordinates are emphasised. Within systems such as Toyota’s kanban system there is no supervisor intervention. Prior plans can be adapted as new information, such as customer demand or defects,
For Japanese workers to participate in such a structure a specific employment incentive system has developed, composed of quasi-permanent employment, a “ranking hierarchy” promotion system and a wage system combining seniority and merit.
Aoki (1988) states that typical conceptions of lifetime employment are exaggerated and not traditional. Instead he prefers to use the term “quasi-permanent employees.” The average Japanese worker does change jobs in their lifetime, just not as frequently as their American counterparts. Also temporary workers are employed in conjunction with the business cycle. Since the mid 1970’s there has been a clear demarcation between regular and temporary workers.
Long tenure particularly desired by J-firms because of the reliance employees’ tacit knowledge in the horizontal information structure. For employees the incentive is linked to a key component of compensation, separation pay. Also for mid-career changers there is an adverse selection problem as employers judge their attributes negatively by leaving mid-career. Consequently job changers have to enter a new company at a lower rank than previously.
As has been mentioned previously, separation pay is a key part of the wage system. Pay consists of a salary, allowances (for housing, for spouse and dependants etc.), merit related bonuses and a lump sum payment at the time of separation. Blue and white collar workers are paid a monthly salary, blurring the status between the two. Only part-timers are paid an hourly wage. Whereas the unionised A-firm has one single wage rate for one job, in the J-firm employees in the same rank may perform different tasks but at the same level of basic pay. J-firms have unions within a single firm or plant rather than within an industry. These so-called enterprise unions do not negotiate wages for each job category, rather they negotiate a base pay for the lowest rank and pay differentials amongst ranks. Overtime and bonuses are related to the business cycle, with the initial reaction to a demand shock being to shorten employees hours rather than lay off workers. If layoffs are necessary senior employers are the most vulnerable, with additional separation payments offered for “early retirement.” 
Contrary to typical perceptions of Japanese firms there is meritocracy. This is through the “ranking hierarchy” in the internal market, and subsequently affects salaries based on rank. Employees compete for promotion throughout their career, with difference more evident in mid-career. Under-performing employees in the 40’s or 50’s may be dispatched to subsidiaries. In 1985 8.2% of employees were dispatched (Aoki 1988, pg. 65). The process of appraisal is an “impartial” process by supervisors based on years of service and merit. It is objective through the check and balances in the relationship between supervisors and subordinates as subordinates can express their grievances through an enterprise union and hinder the supervisor’s progression.
The role of management is to integrate and mediate between a coalition of the shareholders and the employees. This creates a unique corporate structure compared to that of A-firms. In A-firms there is a principal-agent relationship between the shareholders and management. This leads to agency costs and questions about effective monitoring mechanisms. Employees are part of the coalition in J-firms because the separation pay represents “latent financial assets” on the firms accounts (Aoki, 1988).
In the body of shareholders for the J-firm there are three major groups: banks and financial institutions, non-financial business firms and individuals investors. In Japan banks are allowed to hold a maximum of 5% of shares in non-financial firms. One bank acts as a “main bank” to the J-firm with a close relationship to management and a key role in short and long-term credit. The main bank acts as a manager of a loan consortium when groups of banks extend credit-lines to the company. Crucially it has a role as a monitor, with the main bank assuming responsibility for recovery in times of crisis. The main bank has no explicit control in policy making or selection of management, indeed well-run firms are free from invention. Thus the bank’s power is only visible in in poor states. The benefit to this main bank relationship is that a main bank can detect problem quicker. The incentive for management is that they want to avoid outside interference, while for the bank its reputation as an effective and reliable monitor is at stake.
The bank shareholdings are part of a broader range of inter-corporate shareholdings, with two forms of corporate grouping, capital keiretsu and financial keiretsu, emerging.
The screening between high and low productivity workers does exist in the J-firm but its a slow process.
Criticisms based on the applicability of a Aoki’s J-firm theory to all Japanese firms in the real world is an issue that is inherent in formulation of economic theory.
The Nobel Prize winning economist Milton Friedman (1953) in a seminal work, “The methodology of positive economics,” outlined the purpose of economic theory. Foremost, Friedman saw confusion between descriptive accuracy and analytical relevance. To Friedman theories shouldn’t be tested by comparing their assumptions directly with reality. He regarded complete realism as unattainable, rather it is whether a theory’s predictions are good enough for the purpose in hand, whether they are better than the predictions from other theories or whether the theory has the ability for producing new ideas or lines of research. Therefore by following Friedman’s methodology such criticisms are shallow. By Aoki’s own admission his theory is highly stylized and does not represent all J-firms and all A-firms. “Neither a pure J-firm nor a pure A-firm exists” (Aoki 1986, pg. 6). He elaborates noting that U.S. firms are adopting some Japanese practices and vice versa.
Cowling and Tomlinson (2002) generally concur with many of the features posited by Aoki, but by taking a strategic decision-making approach to the keiretsu relations conclude that there are clear similarities between the command structures in both A-firms and J-firms. The see strategic activities as being arranged and controlled at the centre, through the dominance of large firms on their small keiretsu partners. Japan’s production system has been one of mass production with the activities of small firms in the keiretsu network subservient to the requirement of the large firms. Therefore Japan’s large firms have effectively imposed productions decisions, controlled contract conditions and directed technologies and processes on smaller keiretsu partners. The implementation of just-in-time and kaizen has forced suppliers to be adapt production to suit the main contractors. In addition the large core firms use their power from equity holdings and dependency to appoint former executives to key position in suppliers and distributors. For the core firms this allows them to directly communicate and implement corporate strategy in smaller firms. So while there is greater delegation of operational decisions in Japanese firms, this is not the case with strategic decisions. As a result A-firms and J-firms are similar in that strategic decision making is saturated in the elite, corporate hierarchies.
Jackson (2009) in a recent analysis of corporate structure found that Japanese firms seem to a divided into 3 distinct patterns or cluster of corporate structure. The first patterns follows the typical J-firm type expressed in Aoki’s theory. This cluster is consists of large construction, chemical, apparel and textile firms and small firms in machinery and automotive sectors. The second cluster combines market-orientated finance and ownership traits with relational management and employment traits. These firms make heavy use of corporate bonds and feature high levels of foreign investor ownership. They still have a lifetime employment system and high unionisation. This clusters includes Toyota, Hitachi, Canon, NTT Docomo and Mitsubishi group members. The final cluster combined J-firm style finance and corporate governance with market-orientated employment. Unionisation is weak or non-existent in this group, with stock options used frequently for bonuses. This clusters includes trading companies, I.T. service, retail and family-owned firms. Despite some convergence no cluster fully mimics the U.S. style pattern.
From Jackson’s analysis there seems to be clear patterns related to the type of industry a firm is in. This questions the relevance of a national-centric theory of the firm in comparison to an inter-industry theory.
Though there seem to be difference at the micro-micro level these don’t necessarily manifest themselves into difference in industry structure or dominance. Industries that are dominated by larger firms in Japan are also dominated by large firms in America. This could be interpreted as a result of the market forces (Flath 2005)