Writings of Kazuo Nimura

Enterprise Unionism - The Historical Background

1. Introduction

I originally titled this lecture 'A Re-examination of the Debate on Enterprise Unionism', but today I would like to focus on why the organizational pattern of trade unions in post-war Japan differed so much from that in western countries. This has to do with the circumstances that prevailed at the time of the founding of enterprise unions - a central theme in the debate on enterprise unions, and understandably therefore, there has been a great variety of scholarly comment on the matter. I shall first consider the various issues that have emerged from research into 'the origins of enterprise unionism' and then present my own view of these research findings, before finally stating my own position on the central issue.

To start with, I think I ought to make two points clear, although they do not constitute major premises for my argument.

The first is that earlier debate on 'the origins of enterprise unionism' argument made no distinction between 'the basis of the formation' of such unions and 'the basis of their continuation'. However, it seems to me preferable to consider the reasons for the formation of enterprise unions separately from the reasons why, despite various criticisms, such unions have continued for over 40 years. Enterprise unions emerged in a period of political, economic and social upheaval that was unprecedented in Japanese history and which resulted from defeat in war and occupation by the armies of the Allies. There is of course an intimate relationship between the reasons for the formation of such unions and the reasons for their continuation, but I do not believe that the two sets of reasons are identical. I shall look first at the reasons for their emergence.

The second point is that when discussing a particular enterprise union, the pre-war and post-war situations need to be clearly distinguished from each other. If the enterprise union is merely defined as a labor union belonging to a company or workplace, then there were many enterprise unions in existence before the Second World War. There are scholars who focus only on this company-based aspect and who insist that the enterprise union existed before the war, ignoring or playing down differences between those pre-war unions and post-war enterprise unions. Among these scholars are Komatsu Ryūji and Kawanishi Hirosuke, who have both done excellent specialist work on enterprise unionism.

However, I shall maintain that the post-war enterprise union was a different entity from its pre-war counterpart, because the pre-war union was an organization for shop floor, blue collar workers. Furthermore, the pre-war enterprise union did not include all the workers employed by one company. By contrast, the post-war enterprise union was a 'mixed occupation' union that included both blue and white collar workers. It even included the elite cadre of white collar workers who were aiming at or who were marked out for eventual promotion to managerial status. This aspect is almost always ignored in discussion of the formation of enterprise unions, but I feel it deserves more attention.

It is often said that enterprise unions are something peculiarly Japanese, but the All-Japan Seamen's Union (Zen Nippon Kaiin Kumiai) is often cited as an exception. However, when one notes that here too, senior-ranking staff and ordinary crew members belong to the same union, then it too can be regarded as an example of the 'post-war Japanese' union. I believe that such cases are not to be found among foreign seamen's unions. At first sight, even though they belong to the same union, within the organization senior crew members and ordinary crew members are clearly divided. In pre-war Japan too, senior staff belonged to the Seamen's Society (Kaiin Kyōkai) while lower-ranking staff joined the All-Japan Seamen's Union (Nippon Kaiin Kumiai).

The Discovery of the Enterprise Union

I shall now move on to discuss the history of research into the formation of enterprise unions. My starting point, naturally enough, is the discovery that the Japanese labor union, unlike unions in the West, was an enterprise-based organization. The formation of unions began in September 1945, a month after the capitulation. The organization of labor unions, accompanied by 'a storm of strikes', continued through December 1945 and the first half of 1946. The peak was in March 1946, at the time of the passing of the Trade Union Law, when in one month 3,295 unions were formed, and the number of workers in unions increased by 1,030,000. It has often been said that "at the time, no-one noticed that Japanese labor unions were enterprise unions". I myself used to think so, but a little research shows otherwise. Labor leaders had soon realised it already at an early stage, in January 1946. On 17th January, before the formal launch of the Japan Federation of Labor (Sōdōmei), an extended central preparatory committee was formed, and among its resolutions was the following statement:

"Regrettable though it may be that many employees have the wish to form employees' organizations based on individual companies, it is nevertheless a fact. We must of course do away with this delusion, but given the actual situation, we must strive to organize councils of labor representatives working within the same capital ownership structure or councils of workplace delegates that include non-unionized factories; we must organize non-unionized workers and strengthen the (workers') struggle."

This resolution was based on "making core organizational objectives the organization of labor unions at industry level and the formation of a national federation capable of sustaining a strong and solid labor struggle", but it recognized that ordinary workers had "the wish to form employees' organizations based on individual companies".

It would seem that scholars did not at that time take note of this fact. For example, in the journal Rōdō Mondai Kenkyū (Labor Studies) there are a number of papers on the theme of 'the Japaneseness of the Labor Movement', but the term 'enterprise union' was not used at all.
However, apart from labor leaders, there were other people closely observing the labor movement who noticed in 1947 that Japan's labor unions were different from those in the West. For example, the 1946 survey of labor unions included the classificatory terms 'industry-based' and 'occupation-based' but not the term 'enterprise or company-based', yet in surveys from the following year 1947 onwards, 'enterprise or company-based' was at the top of the classification list. Nevertheless, it was not until 1950 that two books appeared which clearly recognized that a characteristic of Japanese unions was that they were based on the individual company and which took up this characteristic. One of these was Sengo rōdōkumiai no jittai (An Investigation of Post-War Labor Unions) published by the Institute of Social Science, University of Tokyo.

The other was Suehiro Izutarō's Nihon rōdōkumiai undōshi (Japanese Trade Unionism:Past and Present).
An Investigation of Post-War Labor Unions was centred on the work of Ōkōchi Kazuo, and surveys carried out by leading labor scholars Ujihara Shōjirō, Shiota Shōbei, and Sumiya Mikio. In August 1947 about 1000 unions had been surveyed and replies received from just 400. In his introduction to the report on this survey, Ōkōchi Kazuo astutely noted that "post-war unions, in a word, share the characteristic of being organizations that are bound up with the company management", and he went on:

Almost all unions were formed at the level of the company and as a result, unions had more the character of employees' unions than workers' unions. Furthermore, very many unions included among their members section chiefs, chief clerk and other such staff in positions of responsibility who are normally considered people who represent the interests of the 'employer'. Apart from special section chiefs, almost 30% of the total number of unions surveyed included ordinary section chiefs, and 50% included staff up to the level of chief clerk. The results of the survey showed that very few post-war labor unions consist of shop floor workers only; most unions, over 80%, are 'mixed occupation unions'. (emphasis Ōkōchi)

As will be understood from this quote, this survey report was the first to clearly demonstrate the characteristics of post-war Japanese trade unions. It can be regarded as a significant milestone of research into Japanese labor studies.
The author of the other work referred to above, Japanese Trade Unionism: Past and Present, Suehiro Izutarō, was the godfather of Japanese labor law, and as the first Chairman of the National Labor Relations Commission, he was both a faithful observer and interpreter of Japan's labor movement before and after the war. Nihon rōdō kumiai undōshi (Japanese Trade Unionism:Past and Present) was originally written so that foreigners could have "an accurate understanding of the links between the history of Japan and its labor unions." The book does not seem to be much read these days, but I believe it to be a work from which one can learn much. It might be exaggerating a little to call it the Japanese equivalent of Sidney and Beatrice Webbs' History of Trade Unionism (1894) but from a vantage point 4-5 years after the beginnings of post-war unionism, it provides a unique description of how the process developed from the pre-war period, and in terms of content, the competence of its excellent observations can only be described as superb. I shall return to this book later, but in its observations that "almost all labor unions were formed in workplace units and on the basis of shared work relationships", and that "it was natural that factory workers and office staff should unite and form a union, and unlike America, where it was normal for high-ranking company officials not to join a union, in Japan they sought to organize themselves in one union together with ordinary employees, and ordinary employees want as many senior employees to join the union as possible", it shows almost the same understanding of the situation as An Investigation of Post-War Labor Unions.

The Migrant Labor Thesis and its Critics

As it became clear that the characteristics of the Japanese labor union differed from those of its counterpart in the West, the question naturally arose as to why. The two books I have referred to of course posited various factors in seeking to answer this question, but the person who brought everything together in the form of 'enterprise unionism theory', who consciously sought to understand the basis of that and became the leader in later research in the field was Ōkōchi Kazuo.

Ōkōchi wrote about 20 books and 200 papers on the theme. Of these, all had a number of basic points in common, but with such a large output, there were inevitably a number of places in which different factors were emphasized. In particular, in his discussion of current events, some of his arguments lack consistency or even contradict each other.
For example, he once advocated that labor leaders should join company managements whereas in another paper he urged labor unionists to abandon enterprise union organization. Rather than a contradiction, however, it should perhaps be said that this showed Ōkōchi's keen sense of realism in responding appropriately to changing circumstances in the movement.

Ōkōchi's first answer to the question of the origins of the enterprise union was his well-known 'migrant labor thesis'. Put simply, this emphasized Japanese workers' tendency to maintain in some way or other their connection with their home village, a characteristic which he designated as a 'migrant labor pattern in a general sense'. In other words, he designated as 'migrant labor in a general sense' those migrant workers who did seasonal work in civil engineering and construction when the fields were fallow, or indentured migrant laborers who worked in cotton mills or silk factories for a fixed period, or else the second and third sons of farming families who worked in factories in the city. He then argued that it was the Japanese characteristics of this waged labor that basically circumscribed Japanese labor issues from the Meiji period (1868-1912) until the Second World War. He went on to explain the origins of the enterprise union in the following way:

'Waged labor is migrant labor' means that industrial areas have difficulty in accumulating the necessary supply of labor. Consequently, labor recruitment occurs on a company basis through agents or through depending on the family and regional ties of owners and of those workers currently employed at the company. In such cases, horizontal labor markets based on occupational categories do not form, and the labor market becomes a vertical one based on the company. Labor unions formed on the basis of such labor markets naturally become enterprise-based organizations.

It is often said that this 'migrant labor thesis' constituted the conventional wisdom in Japanese labor studies for a time, but I do not share this view. Conventional or accepted opinions normally imply that everyone agrees with a particular interpretation and that there are no differing views, but I would say there were few people who supported the 'migrant labor thesis' 100%. Many were critical of it, and various alternative opinions were put forward.
Let us take a look at the critiques of the 'migrant labor thesis' simply in terms of the question of the beginnings of the enterprise union. The first critique came from Ōtomo Yoshio in his book Tōitsuteki rōdōundō no tenbō (Prospects on the Unified Labor Movement) [ Co-written with Endō Shōkichi Funahashi Naomichi, Fujta Wakao and Ōshima Kiyoshi, Rōdō hōritsu junpōsha, 1952] , where he writes:

This explanation cannot satisfactorily account for why the organizational pattern of Japanese labor unions was so different before and after the war. Neither can it adequately explain the fact that there were some post-war unions, albeit only a few, that were not 'company-based employee unions'.

In other words, Ōtomo was pointing out that pre-war Japanese unions such as the Ironworkers' Union in the Meiji period and other organizations since the beginning of the Taishō era (1912-1926) such as the Friendly Society (Yūaikai), the Japan Federation of Labor (Sōdōmei) or the Council of Japanese Labor Unions (Hyōgikai) were mainly occupation- or industry-based unions that transcended the bounds of the single company. In addition, Ōtomo argued that the 'migrant labor thesis' was fatalistic, proceeding from the practical standpoint that it lacked guidelines for action that would enable the union movement to extricate itself from company-based organization. However, among the critical points made by Ōtomo was his inference that the Achilles heel in Ōkōchi's argument was that his 'migrant labor thesis' could not explain the difference between the pre-war and post-war patterns of unionization. I shall discuss later the fact that the subsequent redirection in Ōkōchi's argument aimed at addressing this critique.

The next to criticise the 'migrant labor thesis' was Funahashi Naomichi. In the book Nihon no rōdōkumiai (Japanese Labor Unions, published by Tōyō Keizai Shinpōsha, 1954), which was edited by Ōkōchi, Funahashi developed the following argument:

The organization of labor unions is constrained by so-called objective factors operating in capitalist society such as production processes, industrial structures, structures of authority, and labor markets, and on the other hand, there is the inevitable factor of the workers' class struggle against capital. As a result, labor union organization has an exceedingly empirical and subjective character and in this sense is extremely multifaceted and complex. Without a comprehensive analysis of the relationship between these two factors, it seems reasonable to say that the problem of organization cannot be adequately grasped. Besides the 'objective factors that constrain organization', state regulations, and the nature of economic policies and monopoly have a complicated influence on organization, but the important point is to grasp all these factors in structural terms and to elucidate their relationship to (union) organization. In emphasizing and overestimating a single factor, it is easy to arrive at an erroneous conclusion.

Labor union organizations do not have the same kind of objective social existence as capitalist organizations. Rather, they have to be formed consciously or subjectively by the workers themselves [...] Consequently, by ignoring subjective opportunities for organization and emphasizing only the aspects determined by objective factors, there is the danger of falling into fatalism.

This was a clear criticism of Ōkōchi and of the migrant labor thesis but as no name was cited, it did not spark a debate.
Someone who criticised Ōkōchi from a position quite close to his, getting at Ōkōchi's argument from within, as it were, was Fujita Wakao. His critique did not attract too much attention at the time and was not much taken up, as it was made in a report on an academic conference of the Society for the Study of Social Policy entitled Sengo nihon no rōdōkumiai (Post-war Japanese Trade Unions) and included in the Journal of the Society, Shakaiseisaku gakkai nenpō, vol. 4, (Yūhikaku, 1956). The argument was as follows:

Professor Ōkōchi focuses on the aspect of supply (labor), and in this case we think there are problems with the way in which Professor Ōkōchi deals with this element, but even without entering into that particular discussion, it is not tenable suddenly to argue for company-based union organization only on the basis of the aspect of supply. We too argue from the aspect of supply but we analyze the demand aspect (capital) more than Ōkōchi, believing that without seeking to understand company-based union organization, an adequate grasp of the characteristics of union organization, union management and the union movement is not possible.

The same journal also carried a paper by Fujita titled "Kigyōbetsu kumiai ron to sono hihan ni tsuite"('The Enterprise Union Debate and its Critics'). The paper could be called a review of the debate on studies of enterprise unionism up to the year 1956. It included the following critique of Ōkōchi's position:

The characteristic of Ōkōchi's 'migrant labor thesis' is that it is applied to the whole Japanese labor market before and after the war. Even if this were to be admitted, there were step changes in the structure of the Japanese labor market in both the pre-war and post-war periods. In the post-war period especially, there were times when reforms were being forcibly pushed through and when they were over, functional alterations followed.

'The New Ōkōchi Thesis' and its Critics

There were many other critiques of the 'migrant labor thesis' besides these. Ōkōchi, however, made no response whatsoever to meet all these critiques head-on. Instead, in 1959 he suddenly made a significant shift in his position. This was first made public in April that year at the annual conference of the Society for the Study of Labor History in a report titled "An Historical Examination of the Enterprise Union", a theme that has a very contemporary ring to it. In it, Ōkōchi put forward the following argument:

Before the First World War, the Japanese labor market, especially the market for skilled labor, was horizontal in nature. Workers were not tied to one company and readily changed their place of work. This is why the Ironworkers' Union was a horizontal organization. However, during the economic depression of the late 1920s and early 1930s, major companies succeeded in restricting the horizontal nature of the labor market and adopted labor policies that forced workers to remain with one company. In larger companies a long term employment system was instituted in which workers were hired as 'apprentices' to be trained up within the company and worked until retirement in the same company advancing up a seniority pay scale. Mutual aid systems, company welfare facilities, family allowances, and retirement allowances all served to strengthen and consolidate this modus operandi. Post-war enterprise unions were established on the basis of what had been achieved by locking up the labor market within the company and introducing long-term employment.f

In this argument, the 'migrant labor thesis' that had previously been seen as the ultimate and decisive factor accounting for enterprise unionism was concealed, and in its place, capitalist labor policy emerged as the fundamental factor. The deciding factor was thus changed 180 degree from the characteristics of wage labor to the policies of capital. To distinguish this new direction from the migrant labor thesis, I shall call it 'the new Ōkōchi thesis'.
Since the appearance of this 'new Ōkōchi thesis', there has been little discussion of the origins of enterprise unionism. This is probably because, unlike the migrant labor thesis, the new Ōkōchi thesis' has become the accepted version of events. The focus in Japanese labor history studies has shifted to the study of labor disputes, and the debate about enterprise unionism has moved on from examination of the origins of the phenomenon to issues of seniority in labor-management relations and of internal labor markets; Ōkōchi's ideas are no longer as popular as once they were. However, this is also because the broad structure of the new Ōkōchi thesis, namely, that the former horizontal labor market became fragmented and locked up within companies and that enterprise unions were set up on that basis, was accepted by many people.

I would now like to consider whether this new Ōkōchi thesis can actually explain the emergence of enterprise unionism. As I have mentioned, Ōkōchi radically altered the determining factor 180 degrees from the characteristics of wage labor to the policies of capital. Yet the structure of his argument did not change much, the causal relationships it presents forms a long chain of events, and it shares with his original thesis the fact that it heads up the usual one-way street. The comparison may not be too apt but the structure of the new Ōkōchi thesis reminds one of the proverb kaze ga fukeba okeya ga mōkaru (when the wind blows, barrel-makers profit ); in that sense there is no difference between the migrant labor thesis and the new Ōkōchi thesis. At its core, the argument looks for the origin of the enterprise union in the company-based fragmentation of the labor market. Behind Ōkōchi's argument is the firm conviction that a labor union is nothing more and nothing less than an organization that sells labor. For him, this seems to be a major axiom that requires neither discussion nor proof, a truth valid at all times and in all countries. This is where his view and mine diverge widely. He develops his argument with the assumption that it is a self-evident truth that a labor union is an organization for the sale of wage labor. I believe that to understand Japanese labor unions, it is above all necessary to investigate the historical facts behind the demands and desires on which Japanese workers based their combinations and solidarity. I shall go into my reasons for this later but for now, I should like to examine whether the argument that company-based fragmentation of the labor market was bound to result in labor unions becoming enterprise unions is actually correct and supported by the facts.

I shall argue that, in fact, such a claim is unjustified. I have a number of reasons for this conclusion, but I trust that the following three will be comprehensible:
(1) As I indicated earlier, there were enterprise unions before the war. The earliest was the Japan Railway Society to Correct Abuses (Nittetsu Kyōseikai) which was formed in 1898. However, the key year in which unions were formed one after another was in 1919. Watanabe Tōru published a paper on these developments "Daiichiji Taisen Chokugo no Rōdō Dantai ni tsuite"(Workers' Groups Immediately After World War One) in Jinbun Gakuhō (the Journal of the Institute for Research in Humanities, Kyoto University, Vol. 26, (March 1968) in which he noted that 211 workers' groups were formed in that year alone. Among them were groups that cannot be called labor unions, such as 'female employees supply unions', but of the 211 groups, there were 34 that are classified as 'factory-based groups' (jigyōjōnai dantai). Little is known of the actual conditions of these 34 groups but judging from their names - the Tokyo Arsenal's Koishikawa Rōdōkai, the Osaka Arsenal's Kōjōkai, the Yawata Ironworks' Dōshikai, the Shibaura factory's Giyūkai, the Sumitomo Copper Works' Shinshinkai and the Osaka Ironworks' Teikoku Rōdōsha Kumiai - there were not a few that clearly ought to be called company-based labor unions. Watanabe's classifications also include unions that belonged to organizations that went beyond the single company and which functioned as centers for workers' groups and occupational groups - the Nippon Rōyūkai, which played the central role in the Yawata Ironworks strike of 1919, the Tokyo City Tram Authority's Nippon Kōtsū Rōdō Kumiai, the Dai Nippon Kōzan Rōdō Dōmeikai at Ashio Copper Mine - these were all groups that were actually close to being company-based unions. We have employed the terms company-based union and company-based organization, but most of the workers' groups that appeared in 1919 ought to be called workplace unions rather than company unions. Some were groups that were organized during strikes as autonomous workers' groups; others were groups created by management to oppose them.

It is important to ask whether, in 1919, labor markets in the industries in which these unions were operative - metalworking, engineering, mining - were in fact fragmented and isolated according to company? The answer is a clear "no". Ōkōchi himself said that the reason for this fragmentation and isolation was 'the economic depression of the late 1920s and early 1930s' (e.g. 'Japanese-style Labor-Management Relations and their Antecedents' ("Nihonteki rōshi kankei to sono keifu"), in Ōkōchi Kazuo, Rōshikankeiron no shiteki hatten (The Historical Development of Ideas about Labor-Management Relations)  (Yūhikaku, 1972).

(2) As the question currently under discussion is enterprise unions after the Second World War, it can therefore of course be said that the facts I have just indicated will not suffice to refute the claim that the company-based fragmentation of the labor market gave rise to the company-based union.
Was Japan's labor market then fragmented according to company in 1946-47, when the post-war unions were being organized one after the other? Were the lifetime employment system and the seniority system of real significance at that time? I do not think it can be said that they were at all.

It is hardly necessary to mention that during the Second World War the enormous number of about 7,000,000 young men were conscripted. On the one hand, there was obviously a huge expansion in production for the military - warships, airplanes, and armaments. Naturally enough, there were severe shortages of manpower in factories and mines. The state could not depend only on workers who were new school graduates, and many middle-aged and elderly workers were made to move to industries engaged in production for the military. Every factory had large numbers of conscript laborers, mobilized students, female worker brigades, and Korean and Chinese forced laborers. Then came the surrender. Factories supplying the military closed, and many lost their jobs. Even in those factories that had not done military work, rather than actually producing things, there was recourse to the more profitable practices of 'production sabotage' - siphoning raw materials off into black market channels. Workers, too, abandoned towns and cities where there were food shortages, or quit factories whose prospects looked shaky. If redundancies and voluntary resignations are included, those who lost their jobs are said to have numbered 4,000,000. The 7,000,000 demobbed servicemen, and civilian refugees uprooted from territories overseas swelled the number of unemployed throughout Japan to well over 11,000,000 people. Can it be said that lifetime employment or a company-based fragmentation of the labor market existed and was of any real significance at that time? In a situation in which prices were steadily increasing 10, 20 and 30 times, even any seniority pay lost its ability to keep long-term employees working for one company.

(3) Another important point which must not be overlooked is that the claim that the state of the labor market determined the organizational pattern of labor unions does not explain the fact that many of Japan's post-war unions were of the mixed-occupation type. Ōkōchi solely emphasizes the company-based fragmentation of the labor market, but Japan's labor market was also divided according to educational achievement. The labor markets for shop floor workers and for office staff were distinctly separate. Japan was not the only country in which labor markets for blue collar and white collar workers are completely distinct, but the fragmentation of the Japanese labor market according to educational achievement did not apply only to certain enterprises but throughout society as a whole. Moreover, this education-based fragmentation of the labor market persisted throughout the war and during the chaos of the immediate post-war period. Even today, the labor markets for junior high school, high school and university graduates are completely separate even within the same company. New recruits are divided according to their posting to head office or company workplaces, and their subsequent salary and promotion prospects are completely different. They do of course share the advantages of working for the same company, but there have certainly been many contradictions and conflicts of interest between university-educated white collar staff and middle and high school blue collar staff. Why then do they belong to the same union?

If Ōkōchi's claim were correct - that a labor union is an organization for the sale of labor and thus it is natural for those in the same labor market to belong to the same union - then Japanese unions ought not only to be company-based but also education-based. Furthermore, since, despite the same educational achievement, there are many cases of differing salary and promotion prospects for men and women, it would also be natural for there to be gender-based unions. Yet in reality, there are hardly any such education-based or gender-based unions. This surely implies that there is something wrong with the 'new Ōkōchi thesis' that claims the organizational pattern of labor unions is directly determined by the type of labor market.

The Enterprise Union Arguments of Ōkōchi's Critics

I am sure I shall now be asked what my own views are as to the emergence of the enterprise union and the issue of mixed occupations. However, before I embark on a considered response, as I said earlier, since there were many critics of Ōkōchi's earlier, 'migrant labor' thesis, I would first like to examine these critics' views of the emergence of enterprise unions. I have not time to consider them all, however, nor indeed do I see the need for that, so I shall here restrict myself to one of them, namely, the views of Takahashi Kō. He stresses the need to grasp the fact that the organizational pattern of labor unions is linked to the working class struggle and its various historical conditions and he argues for the need to see that the situation in which the working class was placed immediately after the war made company-based union organization inevitable. With this in mind, he explains the reason why Japanese unions became company-based as follows ("Kigyōbetsu kumiaishugi to sono kokufuku no jōken" (Enterprise Unionism and the Conditions for Overcoming it), in Takahashi Kō, Nihonteki rōshikankei no kenkyū (Studies in Japanese-style Labor-Management Relations) (Miraisha, 1965):

Immediately after the war, the Japanese working class was confronted by the aggressive impacts of unemployment caused by factory closures and mass layoffs, and by substantial cuts in real wages due to inflation. Workers fought back to defend their livelihoods by strikes and 'struggles over production management'. Along with the development of these struggles, the fight to guarantee democratic freedoms, and the Occupation authorities' policy of cultivating labor unions, the number of new unions increased rapidly. Faced with this pressing situation, there was no time for the slow process of persuading individual workers to join a union, and unionization proceeded in leaps and bounds as unions formed at the level of the factory and workplace. On the one hand, capitalist owners recognized that the times were such that the formation of unions could not be avoided and sought to organize them from above, with the company paying all the costs. To combat this tactic of top-down organization, it was also necessary to promote collective union organization in workplace units.

I am in broad agreement with this view. However, I feel that the issue does not relate to the post-war period only but is one that calls for a longer historical perspective. In my view, the pre-war experiences and traditions of the Japanese workers' movement played no small part in the emergence of the company or enterprise union. I shall return to this point later.

I would like to look at another answer given by researchers to the question of why post-war unions became mixed organizations of blue and white collar workers. The book I referred to earlier, An Investigation of Post-War Labor Unions (Sengorōdō no jittai), details the results of a most interesting questionnaire. When asked why they thought unions had become mixed organizations, most workers replied that it was because there was no basic difference between blue and white collar workers. There were also answers to the effect that it was the cooperation between the two groups of workers that made the unions strong, made their campaigns effective and helped industry to recover. In other words, as the bosses were sabotaging production, it was important for office staff and shop floor workers to combine and cooperate to rebuild production and revive industry. Commenting on this questionnaire, Ujihara Shōjiro indicated the following three factors behind mixed occupation unions:

1) during and after the war the livelihoods of blue and white collar workers suffered equally
2) the confusion in structures of authority in the post-war period combined with 1) to remove discrimination between the two groups
3) as the unions' goals included the abolition of systems of status, democratization at work, and participation in management, the ideal campaigning organization was naturally the mixed union.

According to the first point, inflation did away with any real difference between the wages of office staff and shop floor workers, and the livelihoods of both groups were hit equally hard. Also, during the war and in the immediate post-war years production came first, and as there were various kinds of allowances for, and supplies of items to, blue collar workers, real differences in wages between them and white collar workers diminished. In some cases, blue collar workers were even better-off. This was probably a reason behind the formation of mixed unions. However, this alone does not suffice to explain why post-war unions were of the mixed type. It can of course account for why office workers joined the union movement, but even in that case, it would have been possible for office workers, rather than joining mixed unions, to have formed their own white collar organizations and to have campaigned for office staff to get higher wages than shop floor workers, and indeed, such a demand would have been only natural. The same can be said for point number 2 above. It could thus be thought that the third reason - the unions' demand for the removal of status differences and for the democratization of the workplace - is especially significant. However, here too, it can be asked why even white collar workers supported such demands. We must be clear here about the meaning of the demand for the democratization of management under the Occupation. Post-war democracy within the company was, more than anything else, held to mean the demand for the abolition of systems of status. This, in my view, was a decisive factor in the formation of mixed occupation unions. There is a long history to this that goes back well before the post-war years. As this is actually one of my main points today, I shall take it up again at the end.

The Absence of Craft Unions

I now come to my own explanation of the origins of the enterprise union. Actually, my argument is in a sense very simple. It is that in the immediate post-war situation when workers were yet to organize themselves, when they joined forces to press their urgent demands for higher wages, retraction of layoffs, democratization of management and so on, they did so with fellow workers from the same company whom they saw every day at the same workplace. In my view, this was only natural, so natural in fact that it hardly requires explanation. Post-war unions have often been called enterprise-based unions or company-based unions but in the early stages, rather than being enterprise-based, they were in fact factory-based, and many of them were workplace-based organizations. This was pointed out by Takahashi Kō and Tanahashi Taisuke early on, but as people got caught up by the term 'enterprise-based union', it has not received sufficient attention. However, recently, Takagi Ikurō noticed it and pointed out the fact that in the 1960s, the company-wide organizations (kigyōren) that had until then functioned as alliances of workplace-based unions actually became the organizational units of labor unions.

At any rate, it is certain that in the immediate post-war period, the initial organizational pattern of the labor union movement was definitely not, in the case of major companies, that of the company-based union but rather, started out as factory-based, or workplace-based organizations formed by face-to-face groups. Why this was so is, to repeat, in my opinion so natural it hardly justifies examination. I believe that before asking why Japanese labor unions became company-based unions, one ought to ask rather why western labor unions organised themselves beyond the bounds of the company - this is surely the stranger phenomenon.

It may not be a sufficient explanation, but in my view, the reason why workers in Europe and America joined together outside the company or could join together was because they could not imagine doing so without the tradition of the craft guild. It is well-known that labor unions in the West began as occupationally-based craft unions; they were combinations of workers with specialized occupational skills that tightly controlled the training of workers in conduct and job capacities within an apprenticeship system and by this means were able to manage the labor market. They defended their working conditions by restricting the number of apprentices, paying out resettlement allowances and sending workers abroad when there was surplus labor, and by limiting working hours and production capacity. The emergence of such labor unions is unthinkable without the traditions of the craft guilds. It was the existence of these craft unions that gave labor markets in the West their horizontal character. Takahashi Kō drew attention to a significant point in this connection: "One ought not to think that labor markets in the West became horizontal automatically; one also needs to keep in mind the positive aspect that it arose through continuous efforts of the the labor movement to combat the capitalist drive to break up the labor market" (Takahashi, op.cit. p.57). This is an important observation, and keeping it in mind, one sees the significance of the following comment by Suehiro Izutarō in the above-mentioned book Japanese Trade Unionism: Past and Present (Nihon rōdō kumiai undōshi ) :

From the very beginning of Japan's union movement there was no craft union tradition. The idea of skilled workers in the same trade uniting in occupational groups beyond the bounds of the company in which they were employed therefore hardly caught on in post-war Japan either. [...] This is not at all a sign of the health of modern Japanese labor unionism in general; on the contrary, in a sense it can be thought of as the basis of a certain fragility. Out of a need to defend themselves in the face of the dangers of unemployment and the threat to their livelihood caused by the rapid onset of inflation, most Japanese workers in the immediate post-war period simply banded together semi-instinctively and with great speed. Because they normally had little actual experience and knowledge of unions, they knew little about how organizational forms have a deep effect on union activities. It was understandable that workers who did not necessarily share a common interest owing to their different trades should form unions based on the unit of the workplace or the company simply because they shared employment relations, and also understandable that these groups would then go on to band together in industry-wide organizations, but it was not sufficiently realized that when workers from different trades united in one union indiscriminately, this would frequently cause difficulties in cohesion or in union activities.(Nihon rōdō kumiai undōshi (Japanese Trade Unionism: Past and Present) (Nihon rōdō kumiai undōshi kankōkai, 1950, pp.180-181)

I regard Suehiro's indication here that there was no craft unionism tradition in Japan as most important. In response to this, there will be those who point out that there were examples in Japan, such as the Ironworkers' Union, of unions that transcended the individual company, and certainly, in the history of the Japanese labor movement it has been argued that there have been occupationally-based unions, such as the Printers' Union. Ōtomo Yoshio's critique of the migrant labor thesis and Ōkōchi's arguments, insofar as they accepted such critiques, were based on this claim.

However, from research based mainly on documents from branches of the Ironworkers' Union and the Friendly Society (Yūaikai), labor historians have clearly shown over the last decade or so the fact, labor historians have clearly shown over the last decade or so the fact that Japan's labor unions were not occupationally-based unions (craft unions) in the classical western sense. The Ironworkers' Union and the Friendly Society were certainly organizations of workers that went beyond the individual company, but they were not strong enough to keep a grip on the supply of specialized skilled labor nor were they strong enough to control the labor market. They cannot be called craft unions or occupationally-based unions.

Why then were the Ironworkers' Union and others unable to regulate entry into the union through an apprenticeship system? Ikeda Makoto has provided an answer to this question. The following is a summary of his argument.

Japan's heavy industry, which was the organizational base for the Ironworkers' Union, was not something that grew spontaneously, stimulated by the development of light industries. Faced with the military and economic need to tackle the pressure from the advanced countries of the West, it began in the shape of state-sponsored industries such as military arsenals which were developed under government guidance, and large-scale enterprises, such as shipbuilding, which were established under government protection policies. These arsenals and large shipyards imported the latest technology from the West, which made pre-existing technology obsolescent and necessitated training in new skills. There was therefore little technical or human basis for continuity between the traditional artisan class and workers in the new heavy industries. Also, the newly imported technology was rapidly applied to mass production, and trades were increasingly divided up into jobs. At first, there were still workers capable of a variety of manual skills but they did not establish themselves as a stable dominant class. Furthermore, the Sino-Japanese War, the Russo-Japanese War and the First World War spurred the rapid growth of heavy industry, and the demand for labor dramatically increased. Under these circumstances, there was no question of controlling the labor market through the regulation of admission to trades. (Ikeda Makoto, Nihon kikaikō kumiai seiritsu shiron (Nihon hyōronsha, 1970).

This would seem to explain why craft unions did not appear in heavy industry. However, I feel this explanation alone is insufficient to account for why craft unions did not arise in Japan, because after the Meiji Restoration (1868) what counted was the traditional skills of artisans, and there were a number of occupations that fostered technical and human continuity, for example, carpenters, stonemasons, plasterers, roofers, cabinetmakers and other craftsmen associated with the building trade, or metal mineworkers. These trades were not entirely unaffected by the introduction of western technology, but their universal manual skills were not swept away either. Moreover, they had their own autonomous organizations; the carpenters had their taishikō and the miners had the tomokodōmei. If they had had the inclination therefore, it would have been possible for them to have monopolized particular trades through the apprentice system and to have controlled the labor market.

When we think of labor unions, we usually imagine modern factory workers, but when craft unions were the norm in Britain, France and America, the workers who carried the labor union movement were mostly artisans rather than factory workers. For example, the famous Junta - the gatherings of the small number of craft union leaders who dominated the leadership of the British labor movement at the time when unionism was moving from provincial to national organization - mostly consisted of carpenters and joiners. In America, Samuel Gompers, the founder of the Federation of Organized Trades and Labor Unions and the AFL, was a cigar maker, not a modern factory worker. The Knights of Labor included self-employed workers as well as waged workers.
In many countries, the important activists in the early period of labor unionism were craftsmen, especially those in the construction industry or connected to it, but this was not the case in Japan. Carpenters agreed wage levels in the taishikō, and miners who had not served a fixed period of apprenticeship as laid down by the tomokodōmei could not be proper members of the union - such codes of conduct did exist, but they did not have the strength that came from the 'regulatory powers' of the craft union. This is not unrelated to the rapid development of heavy industry in Japan. Even supposing metalworkers had had the kind of codes of conduct that were common in craft unions, they would have opposed the introduction of modern western technology into arsenals and shipyards. In fact, however, hardly any examples are known of organized opposition to such technology. Rather, the foreign experts who came to Japan in the 1850s and 60s were struck by how keen Japanese artisans were to learn how to use the new technology and how easily they adapted to it.

The very lack of so-called 'restrictive practices' among Japanese artisans was one of the preconditions that made the introduction of new technology so easy and the development of Japanese capitalism so rapid. It was not that the new technology and the rapid take-off of Japanese capitalism prevented the emergence of craft unions, but rather, that the absence of craft unions facilitated the speedy transfer of modern technology.

Why then did craft unions not develop in Japan? I am as yet unable to give a clear answer to this question. However, I think it may well be because the character of artisans' organizations in early modern Japanese society and their relation to the authority of the state was very different from the situation in western countries. To put it very concretely, it is because western-style guild regulations did not develop in Japan. Nor can it be asserted that this is necessarily only a Japanese characteristic. For example, according to a British scholar of Russian history, Russian workers before the Revolution were focused on their workplace and thought in terms of the workplace. Rather than the occupational solidarity of turners and finishers, the consciousness of belonging to the same company was stronger; it was close to the feeling of the company as a family. Rather than saying "I am a turner", they would say "I am at the Puchiroff factory". �iSteve Smith, "Craft Consciousness, Class Consciousness: Petrograd 1917", History Workshop Journal no. 11, Spring 1981).

In any discussion of the relation between artisans' guilds and labor unions, the Webbs' critique of Lujo Brentano soon comes to mind. In their critique of his views, the Webbs emphasized that the labor union did not come from the artisans' guild. To the extent that there was no direct link between manual workers' guilds and unions, no 'genealogical' link, the Webbs were probably right. However, what I would like to draw attention to here is the fact that craft unions would never have developed a) without the regulation of admission to the trade enforced by the apprentice system that was devised by the artisans' guilds, b) without the strict assessment of the qualifications of men who had gone through a long seven year apprenticeship and had to pay a high admission fee, and c) without social recognition that it was acceptable to prevent the lowering of working conditions by regulating members' working hours and work efficiency. By 'socially acceptable' is meant not only that artisans demanded these things but that their employers and clients also accepted them.

By contrast, it is thought that Japanese artisans in the Tokugawa period (1603-1867) did not create such social practices - why was this? In passing, I note that guilds were apparently extremely weak bodies in Russia also.

Anger at 'unjust discrimination'

What then was the character of the Japanese labor union? As I have indicated thus far, the Japanese labor union as a seller of labor was a weak organization. Yet Japanese workers themselves did nevertheless sell their labor in an organized fashion. There were frequent negotiations with management in which the strike was used as a weapon. Strikes to press for higher wages or to prevent wage cuts often occurred, even during the period when strikes were effectively banned under the harsh Article 17 of the Public Order and Police Law. However, the fact that working conditions could not be maintained or improved only by recourse to strikes certainly worked against the formation of a horizontal solidarity between workers that went beyond the confines of the single company. Strikes were normally conducted against individual employers, and in such cases, Japanese managers preferred to negotiate only with their own employees and invariably refused mediation from outside bodies. Even today, there are many managements that do not like higher bodies entering collective bargaining processes, but before the war, when workers tried to attract support from other unions, management would absolutely refuse to meet with them.
Not only was this the case with individual managements, government policies also supported such management attitudes. For example, with the aim of "countering joint strike action", Article 17 of the Public Order and Police Law prohibited "tempting or inciting others", but from a certain period onwards, this was not enforced across the board and was used only against activists who were union leaders outside the company. This was done so that if workers inside the company were the leaders, the strike could be suppressed simply by sacking them, but it had the effect of insulating the workers from outside influences. If they wanted to maintain or improve their working conditions, Japanese workers for the most part had no alternative but to strike against individual companies. For unions trying to unite beyond the individual company, this was no small disadvantage. Of course, it was not the case that in particular disputes there were no signs at all of solidarity with other workers beyond the company.

When pre-war labor disputes are analyzed, it becomes only too clear what Japanese workers wanted and expected from their unions. It has often been noted that most Japanese labor disputes were not simply disputes about economic issues; many involved ethical or emotional matters. Especially in large-scale disputes which involved significant violence, the workers' daily frustrations would erupt and showed all the more plainly what they were really concerned about. This was in fact nothing other than 'anger at unjust discrimination'.
For example, the Japan Railways engine drivers' strike of 1898 was not just about a demand for higher wages; there were also demands for engine drivers to be given a status 'equivalent to that of office staff', that their job title be changed from kikankata to kikanshi, that kafu (fireman or stoker) be changed to kikanjoshu (lit. engine assistant), and sōjifu (lit. 'cleaning person') should become kuriinaa (cleaner).
Behind these demands was the fact that engine drivers' wages, which had previously been high compared to those of workers in other occupations, had fallen relatively due to rising prices and rapid pay rises for other workers. However, the important point here was that for them, wages were not just an economic issue or a question of livelihood but were inevitably bound up with 'status equivalent to office staff'. They claimed that they were not just workers but technicians and demanded both that they be treated as office staff and receive higher wages commensurate with that status.

Then there was the example of the Ashio riot of 1907 where the workers showed extreme dissatisfaction with the demands made by the leaders of the Shiseikai (Greater Japan Society of Devotion to Japanese Labor); what the workers wanted was to be sold the same home-grown rice that office staff were eating. At the Ashio copper mine at that time, various items could be bought from the company store and offset against wages. Japanese rice was known as yakuinmai (staff rice) and was only supplied to office staff; ordinary workers could only buy 'Nanking mai' (long grain, imported rice). If they wanted to eat white rice, there was no alternative but to buy it at higher prices from rice merchants in town. Rice was of course not the only issue of discrimination between office workers and mineworkers; others ranged across all aspects of daily living (food, clothing, and shelter), including the size of company accommodation and access to shared or private toilet facilities. Miners (face workers) in particular bitterly resented the bribes that had to be paid to office staff when wages were decided. The demand to be sold white rice symbolized the general anger at discrimination between blue and white collar workers.

What emerged as the focus of pre-war labor disputes was the injustices of managers, and criticism of bosses who practiced unfair discrimination. It goes without saying that this was also the case at Ashio. Another frequent issue in disputes before and after the First World War was the practice of submitting workers to body checks, especially at the entrance to the workplace. In other words, they were being checked under suspicion of possibly having stolen finished items or raw materials. It was felt to be insulting to treat workers like thieves, and the demand was that at least the checks should not be done in public view of everyone."Down with discrimination!" was thus the strong demand of Japanese workers, and the call to "treat us like human beings!", as an important goal of the labor movement, was often heard from the mouths of labor activists. The 'Remains of the 20th Century' column in the Asahi newspaper the other day featured an interview with Yamanouchi Mina, a female activist of the Yūaikai:

Yamauchi Mina often used the expression "an existence lower than the citizens of Tokyo" to describe how female workers were regarded by the rest of society. When female workers passed by in droves, people would say contemptuously "the Tokyo Muslin Company are having a day off today". Yamauchi said in her interview: "At speeches such things were always said. You might think that wages are the main problem for female workers, but the thing that they felt most strongly about was not that". "They want to be treated as proper members of society". Her words were filled with the idea that "female workers are human beings too".

Labor union magazines of the period show that the aim of the labor movement was not just higher wages; the theme of human freedom was always to the fore. Of course, workers would not take at face value the idea that what they were primarily concerned about was not wages but social discrimination. There was no question that wages were an issue. In fact, one can see that social discrimination and low wages were two sides to the same problem. Without really being very conscious of it, we normally say that working conditions, including wages, constitute 'treatment' of workers by employers. The phrases 'improved treatment' and 'higher wages' are normally used interchangeably. Yet, as is clear from phrases such as buchō taigū and kachō taigū ( appropriate behaviour towards the department and section chief respectively), the word taigū means pay and treatment that are closely connected with status. They are, in a sense, symbolic meanings. Whether one's wage is high or low is not just a matter of economics; the amount says something about one's 'value' as a human being.

This thought connects somehow with the point that what many Japanese labor unions were demanding was not just 'higher wages' but an attitude of 'sincerity' or 'good faith' on the part of employers. An often repeated scenario in collective bargaining negotiations is that when the union's demands are rejected or responded to with an offer far below the demand, the Japanese union's response is invariably : "this is a lack of good faith! Give us a more sincere answer!" The amount of the wage hike is, so to speak, taken as a kind of barometer of the sincerity of the other party. Rather than looking to how much the employer can actually afford to pay, the demonstration or non-demonstration of his 'good faith' is what really counts with the union, far beyond what one might expect. A shrewd employer would pay careful attention to the union's demands and after assessing if they were reasonable or not, would appeal for a reduction of the demand on the grounds that the company was not in a position to afford it, offering only a little extra money while making a great play of 'sincerity'. In such cases, the union would, with remarkable placidity, lower the demand. Besides such cases, the issue is connected to a number of other facts, for example, before the war there were some well-known personnel managers in major Japanese companies who practiced something called Mr. So-and-So's management. These included Washio Kageji at Sumitomo, and Fukagawa Masao at Mitsui, who were ardent advocates of 'moral training' and 'spiritual culture'. They stressed the importance of a good understanding with workers but were inclined to rely on emotion rather than rationality. For example, they would go to work earlier than the workforce and clean the toilets, and they did this quietly without making a fuss about it. Most workers are easily influenced by this type of person. A graduate of the law department of Tokyo Imperial University who might well become a company president one day cleaning toilets, which no-one likes to do - the workers would see that as something really admirable.

All the things I have mentioned above show the frustrations harbored and accumulated by Japanese workers, especially before the war, at the social position in which they were placed. There was widespread anger among workers at the unjust ways in which they were treated. This has to do, as might be expected, with the nature of the changes that took place during and after the Meiji Restoration (1868). At that time, the feudal status system collapsed. The new, so-called 'equality of the four classes' (shimin byōdō) was no mere pretence; it had a certain reality. However, in real life, 'equality' in the literal sense of the word was of course not achieved. People chose to work in factories and mines not because they wanted to, but because they were forced to do so by economic hardship, and these jobs were especially looked down on by society as a whole as being 'lower class'. I do not think it is necessary to give examples of this.

Workers were not only discriminated against in society at large but also within management structures. In factories and mines the organisation of production calls for a hierarchy of duties. This is the division of labor required by the production process itself. People who had only had experience of a feudal society did not regard this hierarchy of duties as just a division of labor. It was in a sense only natural that they should understand it in terms of status relationships.
The workers themselves, however, saw it as neither inevitable nor natural. In this sense they differed very much from workers in Britain. It is often said that in the British working class there is the concept of 'them and us'. 'They' are of course the management and their instruments, white collar workers, while 'us' refers to the community of blue collar manual workers. The phrase 'them and us' reflects both the strong sense of solidarity among the working class and the keen sense of hostility against the employers. At the same time, British workers felt it natural that a worker's son should himself be a worker.
In Britain too, of course, there was clear discrimination against workers, and they too bitterly resented it. Yet although both British and Japanese felt resentful of discrimination, there were considerable differences in the way they felt this resentment.

One of these differences was in the way they thought about the range of workers' individual abilities. Workers in the West felt that however strong or skilled an individual worker might be, he should not work too hard or work overtime. If he does that, he will take jobs from his fellows and become a traitor responsible for wage cuts. They thought it natural that one ought to "work slowly so that others don't lose their jobs", whereas in Japan, workers thought it natural that different levels of ability should result in different wage levels. Both groups of workers opposed discrimination, but Japanese workers were not necessarily against discrimination based on merit. This was surely the reason why opposition to piece rate work did not become a general and popular objective in the Japanese labor movement.

Therefore, within the company, if status differences purely reflected individuals' abilities, workers would not have been so exercised about them. However, differentials within companies, especially those between blue and white collar workers, gave rise to a great deal of resentment among blue collar workers. The reason for this I see as academic achievement. Compulsory education became widespread quite early in Japan. The children of landowners and the children of peasants studied side by side at elementary school, and results were judged in accordance with ability and effort. Of course, parental income had some influence on a child's results, but achievement was not determined solely by it; however poor a family might be, they might well have a number of excellent pupils. Yet once the children were through elementary school, parental income increasingly determined the situation. If the family was wealthy, the way was open to middle school, high school and university. Having graduated from high school and university, a young man was naturally a candidate for the 'elite course' office staff in any company. Even if they managed to become office staff, most middle school graduates would remain in lower office ranks; promotion would get them no higher than chief clerk or section chief. If the family was poor, and a child had only an elementary education, he or she would have to go out to work immediately. In such cases, however talented they might be, they would not become office staff. It goes without saying that this situation was especially hard on those children who liked to learn and did well at school. There are many examples of pre-war labor leaders for whom such bitter experiences were the impetus that led them to join the labor movement; this is clear from their autobiographies. All of them tell a tale of great achievement at elementary school, some becoming class monitors or deputy monitors, but although their teachers might have done everything to urge them on to middle school, their family situation did not allow it.

In my view, many of the pre-war labor leaders drew the energy for their activities from such life experiences. One important characteristic of the Japanese labor movement is that within a very short period, the philosophies that underpinned the movement went through one change after another. In the Ironworkers' Union and the early period of the Friendly Society, the basic principle was for workers to raise their social status and gain recognition through the acquisition of skills and the cultivation of character. Yet with the coming of the First World War, the labor movement began to demand that society at large "first recognise each person's humanness" (jinkaku). Cultivation of character was still held to be important for workers, but more than this, there was now the claim that Japanese society itself ought to be reconstructed and reformed. It was this period that gave rise to the strong assertion that the labor movement was a movement not just for higher wages but one that was striving for human freedom and social reconstruction. As methods to facilitate this 'social reconstruction', new philosophies such as guild socialism, syndicalism, and Bolshevism were adopted one after the other. At the annual conference of Sōdōmei (Japan Federation of Labor) in 1922 a fundamental resolution was adopted to the effect that "we affirm that the working class and the capitalist class cannot coexist. With the strength of the union we look forward to the complete liberation of the working class and to the construction of a new society of freedom and equality". This was a totally different line from the organization that had once preached "harmony between capital and labor", compared labor management relations to those between husband and wife and urged the need for mutual understanding between them - and only 10 years had passed between the two stances. This rapid change of outlook has frequently been portrayed as an awakening of the Japanese working class to class consciousness, and that aspect was certainly present; however, why such a rapid turnaround occurred, or rather, could occur was surely because the Friendly Society had been an extremely weak organization in terms of the sale of labor and what had united its members was their anger and frustration at the 'unjust discrimination' against workers in companies and in society at large. I am not convinced that the workers themselves were conscious of such a great difference between demands for "Improvements in social position", "equality and respect for humanness", and "working class liberation and construction of a new society of freedom and equality".

I have come to this conclusion not only because the changes in the philosophy of the labor movement around the First World War can be understood in this light but because I was able to see that the later diffusion of bodies promoting understanding between labor and management that were centered round factory councils in major companies, the emergence of the concept of labor-management unity during the wartime regime and the fact that the enterprise unions after the Second World War became 'mixed occupation unions' were all extensions of this same phenomenon.
The fact that capitalist owners, faced both with the wave of strikes and the sudden emergence of active labor unions in the wake of the First World War, reacted by setting up 'bodies promoting understanding between labor and management' shows, in my view, their grasp of the situation that workers' frustrations were not just focused on wages but "before anything else, the demand to be treated as human beings". Before these bodies were instituted, shop floor workers were expected simply to follow the directions and orders of technicians and office staff, and the institution of these bodies signified management's recognition both that this had been the case and that, since shop floor workers were also human beings, it was natural that they should have various frustrations and demands. Being selected a member of a factory council and being recognized as having the right to sit down with management representatives in the same room and with the opportunity to speak as members of the company's formal structure could not fail to make an impression on workers. Besides the bodies to facilitate good labor-management relations, employers set up company training schools whose graduates were promoted as elite workers. Whereas formerly, only staff who were in close proximity to senior management had received bonuses and pensions, now these were extended to all white collar staff and then, albeit in small amounts, to blue collar workers as well. In addition, monthly salary systems - actually mixed daily and monthly systems - were introduced for shop floor workers. The discriminatory term shokkō (factory hand) was dropped and replaced by various new job titles: kōin and gengyōin (factory worker) kōjin (factory personnel), kōmuin (factory service worker), and ginōsha or gijutsushoku (technician). These new measures testify both to the widespread extent of resentment amongst workers at 'unjust discrimination' and also to the weight that management placed on the recognition of this fact. The fact that Japanese workers' grievances differed from those of the British working class, and the fact that their readiness to accept discrimination on the basis of merit made possible a 'solution' via the means of creating a class of elite workers are points which ought not to be overlooked.

Moreover, attention has also been drawn to the 'Labor and Management Unification Campaign' in the 'Industrial Patriotism' movement (sanpō undō) to promote workers' cooperation in the war effort in the 1930s. These slogans proclaimed that workers and employers were equal in their service to the country. From this viewpoint, workers became 'industrial warriors' (sangyō senshi) and took their place shoulder to shoulder with employers. Individual Industrial Patriotic Associations consisted of the President and the rest of the employees, who were all considered members of the association. What made this possible was clearly the lack of a 'them and us' relationship between shop floor and office staff.

The Occupation authorities' policies of democratization caused a sudden explosion of the workers' long held frustrations. It was only natural that, for the workers, democracy should also mean the demand for the abolition of all status systems within the company. Along with large wage increases, the democratization of management was the main objective of post-war unionism in its initial phase. I would submit that the fact that the unions which emerged in this period were workplace-based organizations and at the same time were mixed occupation unions cannot be properly understood without consideration of the kind of historical background I have sought to elucidate.

[Appendix] In order to put forward my argument as concisely as possible, a number of points had to be omitted. With regard to these, see my paper, "Rōdōsha kaikyū no jōtai to rōdō undō" (Working Class Conditions and the Labor Movement) in Nihon rekishi 18 kindai 5 (The History of Japan vol.18 Modern 5) (Iwanami Kōza, 1975), which has been retitled "Dai-ichiji taisen zengo no rōdō undō to rōshi kankei" (The Labor Movement and Labor-Management Relations Before and After the First World War) and included in this volume.

First published in Japanese in Kenkyū shiryō geppō (Research and Document Monthly), Hōsei University Ōhara Institute for Social Research , no. 305. (March 1984).