UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Postwar industrial relations and the origins of lean production in Japan (1945-1973) Price, John 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1994-894466.pdf [ 6.38MB ]
Metadata
JSON: 831-1.0099231.json
JSON-LD: 831-1.0099231-ld.json
RDF/XML (Pretty): 831-1.0099231-rdf.xml
RDF/JSON: 831-1.0099231-rdf.json
Turtle: 831-1.0099231-turtle.txt
N-Triples: 831-1.0099231-rdf-ntriples.txt
Original Record: 831-1.0099231-source.json
Full Text
831-1.0099231-fulltext.txt
Citation
831-1.0099231.ris

Full Text

Postwar Industrial Relationsand theOrigins of Lean Production in Japan(1945-1973)ByJohn PriceB.A., The University of British Columbia, 1986M.A., The University of British Columbia, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of History)We accept this thesis as conformingto the required standardTHE UNIVITY OYBRITISH COLUMBIA1993@John Price, 1993Signature(s) removed to protect privacyin presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department ofThe University of British ColumbiaVancouver, CanadaDate ‘(a% .9qDE-6 (2/88)Postwar Industrial Relationsand theOrigins of Lean Production in Japan(1945-1973)byJohn PriceAbstractThis thesis examines the evolution of postwar industrial relations in postwarJapan from 1945 to 1973. It analyzes the impact of postwar industrial relationsinstitutions on the origins and development of “lean production” or, as it is otherwiseknown, the Toyota production system. It uses three case studies, Mitsui Coal’s Miikemine in Kyushu, Suzuki Motors in Hamamatsu, and Moriguchi City Hall as anempirical basis for analysis and constructs a schema of industrial relations institutionsthat challenges the conventional “three pillars” interpretation (lifetime employment,seniority-based wages, and enterprise unions).From a historical perspective there were three distinct stages in the evolutionof industrial relations. The first, from 1945-1947 was a labour-dominated periodduring which unions began to develop a distinct factory regime in which they wereequal partners with management and could veto layoffs. Employers rejected thisregime, however, and led an offensive against the independent union movement. Thisoffensive was relatively successful in weakening labour and overturning the newinstitutions, but it engendered further antagonism. Thus the 1950s were characterizedby instability in labour relations and new institutions had to evolve out of theUworkplace. A stable Fordist regime consolidated in the 1960-1973 period.From a comparative perspective and in the context of the development of leanproduction, the author stresses four institutions: tacit and limited job tenure; aperformance-based wage system controlled by management; unions with an enterprise(i.e. market) orientation; and joint consultation. These institutions gave Japaneseindustrial relations their distinctiveness and also help to explain why lean productiondeveloped in Japan.Under the traditional Fordist model, work was broken down into short,repetitive cycles and organized along an assembly line. Employers exerted control bykeeping conceptual activities as their mandate and workers were to simply followinstructions. This study found that work itself did not change substantively under leanproduction but workers participated more in conceptual activities. One of the keyreasons for this was that employers in Japan were able to exercise control not onlythrough the division of labour but through the wage system and enterprise unions aswell. These mechanisms put discrete limits on the scope of worker innovations.They also limited the benefits workers could expect from the system. Lean production represented a new stage in production, identified as lean, intensified Fordism.iliAbstractContentsTablesFiguresContents• . .11• . •v11• . . ixAcknowledgement . . . xGlossaryChapter 1: Introduction . . . Page 1IssuesThe Ubiquitous Three PillarsHistory: Critical and Comparative ApproachesA Specific Comparative FrameworkCase StudiesChapter 2:I.II.III.Iv.Testing the Limits: Workers’Challenge (1945-1948)• . .58• . .65• . .81• . .92.Page 58Chapter 3:I.II.III.Iv.The Japan-U.S. Business Alliance:Capital Retaliates...109...115..134150.Page 108I.II.III.Iv.V...10• .20.40.53Revolution from the TopRevolution at the WorkplaceLabour’s AchievementsShift in AlignmentsPublic Sector AttackedThe Employer OffensiveThe Effects on the Labour MovementAnti-Communism and the Korean WarivChapter 4: Forms and Substance of Labour-ManagementRelationsNikkeiren and the Managerial ArenaResurgence of Independent UnionismSOhyO: Splits and ShuntO• ..163• . .182.202Signs and Sources of ConflictNew Standards in Industrial RelationsMiike Challenges Management Rights • . .248Chapter 6: The Politics of Production in the AutomotiveIndustry: Lean, Intensified Fordism (LIF) Page 258I. The Suzuki RegimeII. Lean, Intensified Fordism: ToyotaIII. Institutional Support for LIFChapter 7: Miike 1960: The Limits of Coercion...315...331• . .339.343.349• . .354...Page 314Chapter 8: High Speed Growth and UnequalDevelopment Page 357Uneven Development: Ascent of the AutomobileIndustryUneven Development: The Decline of CoalThe Public Sector: Moriguchi City HallConsolidation of HegemonyI.II.III.Chapter 5: A System Emerges: Tensions, Limits andChallenges• .Page 160...Page 213I.II.III..216...231• . .261.279.296I. Prelude to Confrontation: Production Politics and theEnergy crisisII. Phase One: Isolation and DivisionIII. Phase Two: Breakaway UnionIV. Phase Three: Nation-Wide MobilizationV. Phase Four: Summer ShowdownVI. CommentaryII.III.IV..362.377.393.408vChapter 9: Conclusion . . . Page 417I. Postwar History and Class Conflict .. .428II. The Workplace Regimes: Case Studies .. .435III. Japan’s Market Hegemony .. .442IV. Some Comparative Assessments .. .465V. Postwar Industrial Relations andLean, Intensified Fordism .. .470Bibliography . . . Page 480viTablesTable 2.1:Table 2.2:Table 3.1:Table 3.2:Table 3.3:Table 3.4:Table 3.5:Table 3.6:Table 3.7:Union Formation, 1945-1948Managerial Councils in 1948Union Members by LabourLaw JurisdictionPrivate Sector LayoffsEmployment Adjustment Standardsfor Regional Public SectorContract Coverage, 1948-1950Decline in Unions and UnionMembership, 1949-1951National Union Affiliations, 1948-1951Firings in Private and Public Sectorduring the Red Purge.82.88...111• ..122• ..123..138..139• ..144..152Grievance Committee Functioningin a Representative SampleManagement Councils and Functionsfrom a Representative SampleNational Personnel Authority Reports andGovernment Implementation,1949-1955Organizational Strength ofthe Miike Local, 1953Union Members as Percentage ofEmployed, 1949-1955Affiliated Membership of UnionCentrals, 1954-1959Average Monthly Hours Worked, byIndustry, 1952-1958Wage Gap in Manufacturing,by Size of Workplace, 1950-1958Unions and Members in Medium and SmallIndustry Organized by SOhyo, 1956-1963Bonus Agreements Negotiated by JCUand Major Coal Operators,1956- 1958Bonuses at Moriguchi Compared toNational Civil Service Employees170.171173.191•..211.217.222• .223.228• .232.234Table 4.1:Table 4.2:Table 4.3:Table 4.4:Table 4.5:Table 5.1:Table 5.2:Table 5.3:Table 5.4:Table 5.5:Table 5.6:viiTable 5.7: Wage Increases at Suzuki, 1954-1960 . . .239Table 5.8: Bonuses at Suzuki, 1954-1959 . . .240Table 5.9: Forms of Union Organization, 1959 . . .244Table 5.10: Male/Female Wage Gap inManufacturing, 1948-1958 . . .248Table 5. 11: Stages in Workplace StruggleMovement at Miike, 1954-1959 . . .250Table 7.1: Replacement Coal Supplies by Source ...336Table 7.2: Nation-wide Mobilization toSupport Miike Union . . .344Table 7.3: Söhyo’s Miike Supporters,by Affiliation . . .345Table 8.1: Fixed Assets and Productivity Levels,Suzuki Motors, 1959-1969 .. .364Table 8.2: Suzuki Plant Expansion, 1960-1971 .. .365Table 8.3: Employment at Suzuki, 1959-1974 . . .366Table 8.4: Suzuki Motors Wage and BonusIncreases, 1960-1975 . . .367Table 8.5: Japan’s Automobile Production andExports, 1960-1973 . . .373Table 8.6: Vehicle Production and Employment . . .374Table 8.7: A Japan-U.S. Comparison of NominalProductivity, 1960-1975 . . .375Table 8.8: Basic Coal IndustryStatistics, 1960-1973 . . .386Table 8.9: Personnel Reductions by the 18 Majors . . .387Table 8.10: Urbanization of Moriguchi, 1960-1972 . . .394Table 8.11: Wage Discrepancies by Age, MoriguchiCity, 1974 .. .402Table 8.12: Moriguch City Employees WageScale, 1974 . . .405Table 8.13: Wages, Labour Productivity andLabour’s Share . . .414viiiFiguresFigure 1.1 Fortune Summary of Production Systems . . .4Figure 1.2 Regulation Theory and Fordism . . .33Figure 1.3 State Intervention and Factory Regimes . . .38Figure 1.4 Location of Case Studies . . .54Figure 2.1 Mitsui’s Miike Facilities . . .66Figure 2.2 Miike Union’s Affiliations . . .70Figure 2.3 The Densan Wage Formula . . .91Figure 9.1 Figure 9.1, Comparison of Labour Relations,Japan-U.S./Canada . . .465Figure 9.2 Union Centralization and Levelof Bargaining .. .466Figure 9.3 Unionization Rates and Level ofBargaining . . .466ixAcknowledgementsMany people have contributed to this study and I sincerely appreciate the help,kindness and advice offered by so many on both sides of the Pacific. First, to myimmediate family, my parents Gwen and Bernie Price, my partner Margaret McGregor, and my children Tommy and Mae, for having put up with a student so “mature”that retirement may arrive before employment.This work would not have been possible without extensive help from manypeople in Japan. Among the many I would like to offer special thanks toHarada Yözö, Iribe Shöji, Kamehara Yoshiaki, KatO Toshio, KOno Kazuo, KubotaTakeshi, Kurumi Yoshiaki, Matsui Susumu, MatsuO Kamachi, Michiyama Fusahito,Nishimura Yasunori, Ogawa KeizO, Sanada Noriaki, Sugita Tomoji, UchiyamaTashiro, Watanabe Akinori, Yamada Shin, and Yamamoto Kikue, all of whom helpedme gain a basic understanding of the workplaces studied here.For their advice and kindness while in Japan, I would like to thank YamamotoKiyoshi, Totsuka Hideo, Hyodö Tsutomu, Hirai YOichi, Matsuzaki Tadashi, Kumazawa Makoto, Watanabe Ben, and Kenmochi Kazumi.On this side of the Pacific, my appreciation to Cohn Gordon and DavidEdgington for their close read of the manuscript and helpful advice. Thanks also toT.G. McGee, Joe Moore, Donald Burton and E. Patricia Tsurumi. And finally, adeep bow to my advisor, William D. Wray, whose patience, attention to detail, andknowledge of Japan’s economic history kept me on track.xGLOSSARYAFL: American Federation of Labor (U.S.).AJPMUF: All Japan Prefectural and Municipal Union Federation (Zen Nihon JichiDantai ROdO Kumiai or JichirO).ClO: Congress of Industrial Organizations (U.S.).CLRB: Central Labour Relations Board (ChuO ROdO Iinkai or ChurOi).TCFTU: International Confederation of Free Trade Unions.ILO: International Labour Organization.JCL: Japan Confederation of Labour (DOmei Kaigi and then DOmei after 1964).JCU: Japan Coalminers’ Union (Nihon Tanko Rodo Kumiai Rengokai or Tanro).JCP: Japan Communist Party (Nihon Kyosan TO).JFL: Japan Federation of Labour (Dai Nihon ROdO SOdOmei or SOdOmei);JPC: Japan Productivity Centre (Nihon Seisan Sei Honbu).JSP: Japan Socialist Party (Nihon Shakai TO).JTUC: Japan Trade Union Congress (Zen Nihon ROdO Kumiai Kaigi or ZenrO).Keidanren: Federation of Economic Organizations (Nihon Keizai Dantai Renmei).LRB: Labour Relations Board (ROdO Iinkai)MMF: Mitsui Mineworkers’ Federation (Mitsui TankO ROdO Kumiai RengOkaior SankOren).MSF: Mitsui Staff Federation (Mitsui TankO Sham Kumiai).NCIU: National Congress of Industrial Unions (Zen Nihon SangyO Betsu ROdOKumiai Kaigi or Sanbetsu)NPA: National Personnel Authority (Jinji In)xiNRWU: National Railway Workers Union (Kokutetsu ROd Kumiai or Kokurö)Nikkeiren: Nihon Keieisha Dantai Renmei (Japan Federation of Employer Organizations).OECD: Organization for Economic Co-operation and Development.SCAP: Supreme Commander of the Allied Powers.SCUF: Satellite City Union Federation, (Eisei TOshi Shokuin ROdO Kumiai RengoKai, or Eitören).TUC: Trade Union Congress (Britain)TUL: Trade Union Law (Nihon ROdO Kumiai HO)UAW: United Auto Workers (United States)WFMCU: Western Federation of Mitsui Coalmine Unions (Nishi Nihon Mitsui TankOROdO Kumiai RengOkai).WFTU: World Federation of Trade UnionsxiiChapter 1IntroductionI. IssuesThis work examines postwar industrial relations and the emergence of a newproduction model in the 1945-1975 period in Japan. The new production prototype,often referred to as lean or flexible production, Toyotaism or management-by-stress,has become the subject of intense scrutiny and debate as researchers, managers andunions attempt to understand the workings and potential impact of the system. As thediverse names assigned the new production model indicate, no common definition orassessment of the system yet exists. But one thing researchers do agree on is that thenew system helped Japan gain an important edge in production quality and efficiency,and on these levels Japan’s production regime has replaced the United States as thestandard-setter internationally. This has ushered in a new era of learning from Japan,particularly in the fields of industrial relations and production management.A New Industrial Relations Model?In 1975, an O.E.C.D. (Organization for Economic Co-operation and Development) mission to Japan concluded that: “Though the Japanese industrial relationssystem seems remarkably well adapted to the functional needs of a democraticmarket-economy, the cultural differences between Japan and other industrial countriesIntroduction. . .1are such that it is unlikely that any particular feature could be extracted for emulationoutside Japan.”1 A Canadian government commissioned report by industrial relationsconsultant C. Connaghan reached similar conclusions as late as 1982.2Yet, within the decade, researchers in Canada and abroad began to advocatedramatically contrary conclusions. The 1985 Canadian MacDonald RoyalCommission on the Economy concluded: “Commissioners believe there isconsiderable potential for, and considerable advantages to be gained from, the morewidespread use of certain features (such as joint consultation) of the Japanese systemin Canada.”3 Some researchers in Japan had, at nearly the same time, also noted thechange in orientation. As Shimada Haruo put it: “The new focus is on trying todistinguish Japan’s own logic of development in industrial relations, with the intent ofidentifying elements possibly transferable to Western advanced nations. This trendmay be described as a ‘search for a new general model of industrial relations throughJapanese experience.”41 Organisation for Economic Co-operation and Development, TheDevelopment of Industrial Relations Systems: Some Implications of theJapanese Experience, (Paris, O.E.C.D., 1977), p. 40.2 Charles J. Connaghan, The Japanese Way, Contemporary IndustrialRelations, (Ottawa, Labour Canada, 1982). The Royal Commission on the State of the Economic Union, (Ottawa,Supplies and Services Canada, 1985), p. 713.‘. Shimada Haruo, “Japanese Industrial Relations--A New General Model?A Survey of the English-language Literature,” in T. Shirai ed.,Contemporary Industrial Relations in Japan, (Madison, University ofWisconsin Press, 1983), p. 25. Shimada correctly points out that thechange in interpretation is closely related to Japan’s economic successand the corresponding decline of the United States. This has implied adecline in what has been hitherto known as the convergence thesis. Thisthesis, most explicitly developed by Clark Kerr and others, held that asstates become modernized the industrial relations systems would becomeincreasingly similar to that of the United States (see G. Bamber & R.Lansbury, “Studying International and Comparative Industrial Relations,”Introduction. . .2Thus in less than a decade labour relations in Japan jumped from beingunique, quaint, and culture-bound to stand as a potential general model--modern,egalitarian, and transferable at the same time! Indeed, the triumph of the NewPrometheus--Japan--has had a profound impact on Western perceptions of Japan’ssocial institutions. Is this new interpretation of things Japanese founded on anydeeper understanding than earlier perceptions?Lean Production: Post-Fordism?The trend to emulate Japan extends beyond the sphere of labour relationsnarrowly defined. More than a decade ago, companies in North America began tointroduce quality circles as a panacea for a perceived decline in the rate ofproductivity improvement. Today, what was once a trend has mushroomed into afull-scale movement to promote production systems developed in Japan’s largefactories. Fortune magazine recently summarized its perception of the advantages tobe gained from the “lean/flexible” system and their assessment is reproduced inFigure 1.1.As the Fortune illustration indicates, managers in North America are closelyin their book, International and Comparative Industrial Relations,(London, Allen & Unwin, 1987) for details on the origins andtransformations in convergence theory) . Many researchers have, along withShimada, breathed a sigh of relief at the decline of the convergencetheory. Yet, given Shimada’s own conclusions about a ‘new general model’one can not escape the nagging feeling that, rather than being a deadletter, the convergence thesis is simply being rerouted towards the newPrometheus.. Many names have been assigned to the model: lean production,innovation-mediated production, management-by-stress, post-Fordism andflexible production are among the most well known.Introduction. . .3Figure 1.1: Fortune Summary of Production SystemsTWO WAYS OF MAKING THINGSThe Lean/Flexible System The Buffered/Rigid System(New Japanese Style) (Traditional American Style)Can be profitable making small batches of products Profitable only when making large batches.The product and process form making it are The process is designed after the product has beendesigned concurrently. designed.The lean inventory turns over fast. The fat inventory turns over slowly.Suppliers are helped, informed and kept close. Suppliers are kept at arm’s length.Engineers search widely for ideas and technology. Engineers are insular, don’t welcome outside ideas.Employees learn several skills, work well in teams. Employees are compartmentalized.The company stresses continuous small improve- The company looks for the big breakthroughs.ments.The customers’ orders pull the products through The system pushed products through to thethe factory. customers.Source: Fortune, “Manufacturing the Right Way”, (May 21, 1 990), p. 60.examining the apparent success of Japan’s production methods. Academics such asDaniel Roos, director of MIT’s prestigious five-year International Motor VehicleProgram, have concluded that lean production will be the wave of the future and thatToyota was the birthplace of this new production model.6 Others, such as MichaelPiore and Charles Sabel, posit that the world is moving towards another industrialrevolution based on flexible specialization.7 The concept of a new mode of. James P. Womack, Daniel T. Jones & Daniel Roos, The Machine thatChanged the World, (New York, Macmillan, 1990). Michael Piore & Charles Sabel, The Second Industrial Divide:Possibilities for Prosperity, (New York, Basic Books, 1984) . MartinKenney and Richard Florida disparage Piore and Sabel’s work as “fablesbased on northern Italy,” in their recent volume, Beyond Mass Production,(Princeton, Princeton University Press, 1993) p. 301 and pp. 12-13. Themost extensive, early review of the literature on flexibility and Fordismis Stephen Wood’s introductory essay in the work he edited, TheTransformation of Work?, (London, Unwin Hyman, 1989)Introduction. . .4production has thus inspired futurology of the academic variety but it has also inspiredpopular versions along the same theme. Consultants such as Tom Peters (In Searchof Excellence) and Alvin Toffler (Powershift) are also clearing the way for a changein production philosophy.8,9Support for the concept of a new production paradigm encompasses a diversityseldom seen. Writers in Fortune Magazine, for example, have been joined by groupssuch as the British Communist Party in supporting the new production model.1°Despite interpretive differences, this phalanx of opinion heralds what some contend isa fundamental paradigm shift in production systems to a ‘post-Fordist’ model.Central to the post-Fordist thesis is that the Japanese production system, includinglabour relations, has led us into a new production era. Among the most eloquent andsophisticated advocates of the post-Fordist thesis are Martin Kenney and RichardFlorida. They have argued that Japan’s innovative production regime “replaces thetask fragmentation, functional specialization, mechanization, and assembly-lineprinciples of Fordism with a social organization of production based on work teams,job rotation, learning by doing, flexible production, and integrated production8 Tom Peters, Thriving on Chaos, (New York, Harper and Row, 1987). Alvin Toffler, Powershift, (New York, Bantam, 1990)10 For the former see Robin Murray, “Life After Henry (Ford)” inMarxism Today, (October, 1988) . For the perspective of Americanmanufacturers see Fortune’s special coverage on “Manufacturing the RightWay” in its May 21, 1990 edition. This position is also articulated (withvariations of course) in Tom Peters, Thriving on Chaos, (New York, Harper& Row, 1987) and in Womack et al, The Machine that Changed the World (NewYork, Macmillan, 1990)Introduction. . .5complexes.”Yet, voices of dissent can be detected.From New Zealand, KatO and Steven launched a scathing critique of theportrayal of Japan as a new stage in capitalist development. According to them,Japanese management represents not a higher form of capitalism but rather more‘primitive forms of social control” which they associate with Reaganism and Thatcherism.’2 Others who have criticized the post-Fordist thesis include Dohse, Jurgens& Malsch from Germany; Sheila Rowbotham in Great Britain; Mike Parker and JaneSlaughter in the United States and, most recently, Christian Berggren of Sweden.’3Kenney and Florida’s latest work, Beyond Mass Production, represents themost thorough and sophisticated articulation of the post-Fordist thesis. They havenow dubbed the new production paradigm “innovation-mediated production” which,according to the authors, will define “the future of the advanced capitalist world.”4Although most of their latest volume examines the transfer and diffusion of the newmodel, it does contain a substantive chapter that examines the genesis of the system in“. Martin Kenney and Richard Florida, “Beyond Mass Production:Production and the Labor Process in Japan” in Politics and Society, (Vol.16 No. 1, March 1988), p. 122.12 Kate Tetsur & Rob Steven, “Is Japanese Capitalism “PostFordist”?, a paper presented to the 8th New Zealand Asian StudiesConference, Christchurch, August 17-19, 1989.13 See Knuth Dohse, Ulrich Jurgens & Thomas Malsch, “From ‘Fordism’to ‘Toyotaism’? The Social Organization of the Labor Process in theJapanese Automobile Industry” in Politics and Society, (Vol. 14, No. 2,1985), pp. 115-146; Sheila Rowbotham, “Post-Fordism” in Z Magazine,(September 1990); Mike Parker and Jane Slaughter, Choosing Sides: Unionsand the Team Concept (Boston, South End Press, 1988); Christian Berggren,Alternatives to Lean Production, (Ithaca, ILR Press, 1992)14 Kenney & Florida, Beyond Mass Production: The Japanese System andIts Transfer to the US, (New York, Oxford University Press, 1993), p. 316.Introduction. . .6Japan.In Beyond Mass Production (Chapter 2, “Beyond Fordism”), the authors havemade an effort to provide a historical perspective that links industrial relations toproduction management:This chapter outlines such a theory [the dynamics of Japanesecapitalism] by exploring the origins, historical determinants, andevolution of innovation-mediated production in Japan. The basiccontours of the argument are as follows. The rise of innovation-mediated production in Japan was tied in large measure to thespecific constellation of political and economic forces acting onJapan in the immediate postwar years. During this crucial period,intense industrial unrest at the point of production, popularstruggle, and class conflict unleased a set of forces that altered thebalance of class power or “class accord,” produced a distinctpattern of capital-labor accommodation, and resulted in a dramaticrestructuring of work and production organization. 15Key in this early period were the struggles for production control (1946-47) which,according to Kenney and Florida, “essentially, established the roots of the Japanesesystem of team-based work organization.”16 The authors recognize that the early‘labourist’ period was superseded by a managerial offensive in 1949, but they contendthat “many of the characteristics now interpreted as indicating capital’s control oflabor were initially labor demands. Like the postwar accords of the United States andWestern European countries, only later were these demands integrated into the logicof capitalist accumulation.”’7 This postwar accommodation was reflected in the newsystem of industrial relations that “revolved around guaranteed long-term15 Kenney & Florida, Beyond Mass Production, pp. 23-24.16 Ibid., p. 28.Kenney & Florida, Beyond Mass Production, p. 29.Introduction. . .7employment, a seniority-based wage system, and enterprise unionism for the core ofthe labor force.”I share Kennedy and Florida’s assessment of Japan’s importance and inattempting to present a cohesive portrait of the origins of the system, a portrait thattakes into account workers and their struggles, they have provided a thoughtful andreasoned interpretation. It is an interpretation, however, which I do not entirelyshare. In their effort to portray the new production paradigm as the result of aspecial accommodation between labour and management, Kenney and Florida areobliged to distort the dynamics of postwar conflict in Japan, and end up relying onwhat I consider a caricature of the industrial relations system, the alleged three pillarsof job tenure, seniority based wages and enterprise unions, to justify their views. Onthe other hand, Kenney and Florida quite rightly point out that critical studies caneasily fall into the trap of portraying Japan’s development as a return to the despotismof coercive capitalism, or of dwelling on the theme of super-exploitation.The extreme polarization in the debate represents a general difficulty incoming to terms with what I call the paradox of Japan’s production politics: On onelevel, workers are very much involved in production, work in teams, and rotate jobs;yet, on another level the system maintains a strong bias towards mass production andexploitation. This study acknowledges that both aspects did develop as integral partsof the system and attempts to explore the paradox, not dismiss it.The research from this study reveals that, indeed, one of the main features ofthe industrial relations system that evolved in postwar Japan was the extensive controlIntroduction. . .8employers enjoyed over work and workers. This control has never gone unchallengedand some unions succeeded in modifying the system. They were a minority. Ingeneral, the postwar compromise on the level of the shop floor ended up a lop-sidedaffair with employers enjoying extensive control. This control did not, however, leadto the despotic regimes of early capitalism, although during the 1950s there was thattendency. Instead, the regime evolved into a variant of a Fordist system, similar inits high productivity-high wage formula to Fordist regimes in other countries but withsome additional attributes.However, Japan’s production system retained the stamp of extensive employercontrol and the hegemonic regime that developed was unique in that it allowed forextensive employee involvement without altering the norms of Fordist production.Instead, the regime shaped the input that workers had in production matters and, Iwill argue, it undermined workers’ ability to articulate their own independent agendafor the workplace and this, in the end, diminished labour’s capacity to extract thebenefits one might have expected from such an efficient system. Thus Japan’scontemporary factory regime could not break with Fordism, and in the automobileindustry at least, the assembly line, repetitive and routine jobs, and standardizationremained at the heart of the production process. But it was a dynamic system, onethat continually renewed itself in response to competitive pressures. I refer to thisdynamic system as lean, intensified Fordism and it represents a higher stage in theevolution of capitalist productivity, one that must be studied, and learned from. Todo so, however, requires that we return to the origins of lean production, that weIntroduction.. .9come to terms with the paradox of Japan’s production politics by re-examining thehistorical relationship between industrial relations and production norms as theyevolved in postwar Japan.II. The Ubiquitous Three PillarsContemporary fascination with production in Japan, spreading as rapidly asJapan’s automobile assembly plants have moved abroad, has unfortunately tendedtoward the superficial and the short-sighted. Although there is a new appreciation ofJapan’s economic strength, our perception of the reasons for that strength remainsbased on old interpretations. Nowhere is this more true than in the field of industrialrelations. The conventional “three pillars” interpretation of Japan’s labour-management relations, adopted in a modified form by Kenney and Florida, has tendedto perpetuate a stereotype of workers in Japan as loyal employees bound to thecompany through paternalistic employment practices of cradle-to-grave employment,wages that increase with seniority, and unions that are focused on the enterprise.Although not entirely lacking in substance, this portrait of industrial relations in Japanis more caricature than real. Furthermore, as a generalization it leaves little room fordigression. The three pillars typology, in the end, prevents us from exploring theintricate web of Japan’s work life and, when integrated as part of the new productionmodel, promotes a vision for workers that, from a comparative perspective, cannot bejustified. It is understandable that policy analysts and overseas commentators,when looking to understand the new factory order, would quickly refer to the threeIntroduction. . .10pillars theory of Japan’s industrial relations. That the system is supposedly structuredaround the three pillars of lifetime employment, seniority-based wages and enterpriseunionism has become conventional wisdom. For the past thirty years, it has beenoffered up as standard fare with few exceptions. And, as we shall see, even recenthistoriography has been unable to break the iron mould of the three pillars.James Abegglen was among the first of many scholars and industrial relationsspecialists to elaborate the three pillars interpretation of Japan’s industrial relations.In his 1958 classic, The Japanese Factory, he postulated that jobs were permanent andthat wages were based on seniority.’8 Japanese employers themselves began topromote a similar interpretation to Abegglen. In 1963, Sakurada Takeshi,representing Nikkeiren (Federation of Employer Organizations), used the Japaneseterm “three golden treasures” to describe the main features of the labour relationssystem. Since the 1960s, citing the three pillars has become de rigueur for anyaccount of labour relations. Astonishingly, adherence to these features transcendsphases of interpretations, academic fields and even the political spectrum. WhetherOECD reports or leftish journals in the United States, the three pillars apparentlyprovides support for every interpretive bent.’918 James Abegglen, The Japanese Factory: Aspects of its SocialOrganization, (Glencoe, The Free Press, 1958)In Shimada’s literature survey he posits that the three pillarsterminology is derived from the Japanese term ‘three golden treasures’.This is analogous to the three treasures (jewels, a mirror, and a sword)bestowed upon the gods Izanami and Izanagi according to folk legend.Sakurada Takeshi used the same terms in reference to lifetime employment,seniority-based wages and enterprise unionism. On the left, see DavidLevine’s article “Japan’s Other Export” in Dollars and Sense, (September,1990), pp. 17-24.Introduction. . .11Of all the works in English based on the three pillars interpretation none hasbeen more influential than that of Ronald Dore, the eminent British sociologist. Inhis classic work, British Factory-Japanese Factory, Dore offered the most sophisticated schema of the three pillars theory as well as a treatise on Japan’s labourrelations history. He hypothesized that, in the prewar period, Japan’s employerspossessed a dual character--on the one hand they, like their counterparts in othercountries, had a fundamental interest in the market (profits, expansion and efficiency)but that this was blended with Confucian benevolence which gave rise to the prewar“firm as extended family” pattern of industrial organization. Under this constructionemployers, although authoritarian, were paternalistic, according their employeescertain welfare benefits, periodic bonuses, and so forth. Furthermore, employerswere convinced of the necessity of working with unions from 1922 on, and were ableto shape union-management relations before unions became too strong. Thus Japan,according to this cultural interpretation, was able to avoid the fate of Britain wherethe early and drawn out process led to acute class conflicts and “the antiqueinflexibility of her trade union institutions.”20In the postwar period, according to Dore, Japan underwent a social-democraticrevolution immediately after the war and employers unquestionably accepted unionsand abolished the statuses between staff and manual workers. This gave rise to thepostwar “enterprise-as-community” pattern of industrial relations based on the threepillars of lifetime employment, seniority wages and enterprise unionism. As a20 Ronald Dore, British Factory-Japanese Factory, (Berkeley,University of California Press), p. 420.Introduction. . .12consequence of this “late development” syndrome, industrial relations in Japan haveleapfrogged ahead of Britain on the road to “democratic corporatism.”If anything, Dore’s recent publications, Flexible Rigidities (Athione Press,1986), and Taking Japan Seriously (Stanford University Press, 1987) reinforce hisinitial contentions and constitute aggressive advocacy of the Japanese model as heinterprets it.21 Due to Japan’s Confucian roots, according to Dore, the ‘firm-as-community’ has given employees an equal if not superior footing with managers andshareholders, enterprise proceeds are fairly divided and decision-making is from thebottom up. In comparative terms, Dore associates his ‘firm-as-community’ modelwith Swedish social-democratic institutions and sees in it the future direction ofinternational worklife.22 Although not couched in the same terms, Dore in factprojects a post-Fordist vision of Japan similar to that espoused by Kenney andFlorida.To be fair, in the past decade a number of scholars haven taken issue with thethree pillars paradigm or at least cautioned us about its limits. Koike, for example,has directly challenged the concept of permanent employment.23 And, in anextensive survey of English-language literature on Japan’s industrial relations,21 The full title of Dore’s most recent work is Taking JapanSeriously, A Confucian Perspective on Leading Economic Issues (emphasisadded) Dore explicitly re-affirms his adherence to the three pillarsparadigm on page 9.22 Dore’s works represent only the sophisticated cutting edge to whathas become a deluge of materials that advocate Japanese-style institutionsbe it in production systems, education or labour relations.23 Koike Kazuo, Understanding Industrial Relations in Modern Japan,(London, Macmillan, 1988)Introduction. .. 13Shimada warned that the three pillars stereotype “tends to overshadow facts that donot conform with it and to discourage alternative interpretations.”24 Yet Shimadahimself seems to forget his own admonitions when it comes to dealing with one of thepillars, so-called ‘enterprise unions.’For Japanese workers the enterprise union was the only, and mostnatural, form of organization because their basic common interestas industrial workers had been formulated within an individualenterprise.25Not only does he continue to uphold the validity of the enterprise union model, he hasassailed any critique of enterprise unions:An interesting example is Galenson (1976). He analyzed theoperation of the Japanese labour market and industrial relationssystem and concluded in effect, that Japanese unions have notgenerated the strength necessary to represent workers’ demandsproperly or to protect their interests, and that they have failed tosecure the workers’ due share of the gains from economic growth.A view of this kind apparently assumes that American or Anglo-Saxon trade unionism is almost the sole ideal type and dismissessome differing but important attributes that make Japanese-typeunions effective.26This vicious cycle of pointing out limits only to reinforce them illustrates theunderlying persistence of the three pillars typology, not to mention the nationalistpitfalls of comparing and evaluating differing labour relations institutions. EvenAndrew Gordon, despite the historical insights in his book, The Evolution of Labour24 Shimada Haruo, “Japanese Industrial Relations--A New GeneralModel?”, p. 10.25 Shimada Haruo & Shirai Taishiro, “Japan” in Dunlop and Galensoneds., Labor in the Twentieth Century, (New York, Academic Press, 1978), p.258 [emphasis added]26 Shimada Haruo, “Japanese Industrial Relations--A New GeneralModel?”, p. 8.Introduction. . .14Relations in Japan, accepted the three pillars framework, thereby weakening hiswork’s potential significance. Nevertheless, Gordon’s study constituted a benchmarkin English-language scholarship on Japan’s labour history and deserves furthercomment. Furthermore, Kenney and Florida base much of their analysis of Japan’sindustrial relations on Gordon.27Gordon’s study attempted to discover the nature and origins of industrialrelations patterns in Japan through case studies of a number of firms in heavyindustry. Its tremendous strength is derived from the fact it traced patterns ofindustrial relations at these firms over the course of a whole century; it emphasizedthe role of workers’ struggles in shaping the labour-management relationship; and itrefused to be bound by convergence theories. Gordon concluded that there weremany aspects of continuity in industrial relations institutions (the bonus system forexample) from the pre to the postwar period. He also posited that for men in largeenterprises, “the postwar settlement emerges as a far-reaching transformation of thelabour relationship.”28 Male workers in large enterprises were finally given‘membership’ into the enterprise community and given the benefits of the threepillars. However, he cautions against attributing too much importance to, or pittingconflict against, culture. Both workers and managers manipulated culture to promote27 Andrew Gordon, The Evolution of Labor Relations in Japan: HeavyIndustry, 1853-1955, (Cambridge, Harvard University Press, 1985) . Kenneyand Florida base their interpretation of postwar labour relations in Japanon Gordon although it is not clear if Gordon agrees with the particularspin they have given his study. See their acknowledgment in Beyond MassProduction, p. 27.28 Andrew Gordon, The Evolution of Labour Relations in Japan, p. 411.Introduction. . .15their own interests.Gordon’s study had its own specific framework and its limitations. Forexample, his conclusion that the postwar settlement represented a significanttransformation of labour relations was specifically relative to the pre-war and wartimeregimes. He also concluded that:Workers did become part of the organization to a far greater extentthan before or during World War II. Although managers rejectedtheir program of control, participation, contractually secure jobs,and explicit livelihood wages, they conceded the status of‘employee,’ the respect and security of a monthly wage, and theright to use all facilities to an expanding pooi of workers. Andthey worked out an implicit system of job security and livelihoodwages acceptable to most employees. From the perspective of thelate twentieth century, this may look like a cheap set ofconcessions, largely symbolic, often imposed from above, andactually in management interest. But remember how different thesituation had been in the 1930s and during the war.29In other words, Gordon’s conclusions were strictly in relationship to Japan’s domesticevolution. Today, when Japan’s industrial relations and production processes arebeing cast as a potential model for other countries, such conclusions must be reassessed from a comparative perspective. In such a light, the weakness of the threepillars theory as a framework of analysis becomes much more apparent.Although he adopted the three pillars analogy, Gordon, writing a decade ago,was also aware of the potential risks associated with invoking such a typology.Regarding jobs, for example, Gordon concluded that the term permanent employment29 Andrew Gordon, The Evolution of Labour Relations in Japan, p. 411.Introduction... 16or lifetime jobs was misleading.3° He also acknowledged that wages were contingenton much more than seniority. Yet despite these qualifications he continued to acceptthe designation of the wage system as seniority-based. Furthermore, Gordon neverseriously explored the enterprise union pillar, and by equating enterprise unions with“company unions,” perpetuated confusion.3’ The greatest limitation of his work as abasis for understanding postwar Japan, however, is that he really examined postwardevelopments only in a single chapter. His exclusive emphasis on emphasis on heavyindustry also narrowed the scope of his postwar investigation. And the idea thatJapan’s industrial relations system consolidated between “the late 1940s and the mid1950s.. .to endure relatively unchanged for at least three decades,” obliges Gordon toomit such institutions as annual bargaining from his narrative and undermined theconstruction of an appropriate periodization for the postwar era.The present study, however, challenges the three pillars interpretation ofJapan’s industrial relations as well as Kenney and Florida’s general post-Fordistthesis. Where they stress continuity between the early labour triumphs and the post-1949 period, I found discontinuity and qualitative changes, particularly in the natureof the wage system, union orientation, and union input over job levels. In fact, indefining the contours of postwar industrial relations, this study will fundamentallychallenge the conventional wisdom about Japan’s industrial relations. It will arguethat not only does the three pillar interpretation have limited scope (most30 Andrew Gordon, The Evolution of Labour Relations in Japan, p. 2.‘. Ibid., p. 3.Introduction.. .17commentators now acknowledge that employees enjoying permanent employmentstatus represent less than 30 percent of the work force), but even within the limitedconfines of large enterprises the three pillars was and remains an inaccurate andinadequate description of the institutions of employer-employee relations.In re-exploring the history of postwar labour relations, it became clear that thedominant pattern of industrial relations in Japan reflected greater employer dominationand, consequentially, greater market influence in shaping workplace values than inother industrialized countries. This was the result of labour’s defeat in intense classstruggle in the early postwar period, a defeat which led not to an accommodation oflabour’s demands but in fact to their reversal and to long term weakening of labour’sability to shape the Fordist compromise that later emerged in the 1950s and 1960s.To be sure, some of the features of the later accommodation, such as jointconsultation committees, harkened back to the management councils of the 1946-1948period in which labour had representative equity and exercised a right to veto layoffs.But joint consultation committees were a pale imitation of the former managementcouncils, and given the overall relation of forces between labour and management, didnot resemble, for example, the forms of co-determination that emerged in continentalEurope in the postwar period.Instead, the system of industrial relations that emerged reflected the impact ofthis strong employer influence and undermined workers’ capacity to remove theiremployment conditions from the competitive market and employer control. Takingone specific area as example, wage determination in Japan’s workplaces becameIntroduction. . .18subject to employer control through an extensive system of regular personnelevaluations. In other words, instead of wages being calculated on the basis of strictobjective criteria (age or accumulated service or jobs) they became the subject ofunilateral management evaluation of a worker’s worth. Other researchers, includingGordon and Dore, have noted this point, but few have stressed how widespread thisprocedure had become and even fewer have understood or analyzed the fundamentalimplications of this type of wage determination. Instead we have been indulged withthe constant refrain of the “seniority-based” character of the wage system in Japan.The ability of employers to manipulate the industrial relations environment totheir advantage gave them some important advantages in maximizing capitalaccumulation. I will argue that Japan’s Fordism, while allowing for higher wages,achieved its rapid development partially through a degree of exploitation of labourgreater than in other industrialized countries. As a result of extensive control in thework place, employers were able to use labour more flexibly than in automobileplants in North America. But this approach also exacted a harsh toll from the labourforce.Furthermore, because the industrial relations environment differed from, forexample, that of the United States or Canada, the types of organizational innovationsthat occurred in the automobile industry displayed important variations from Fordistnorms in the U.S. I will argue, however, that these variations did not provoke afundamental break with Fordist norms of work organization. Instead, what evolved inthe automobile industry at least, was a leaner, more intense version of Fordism. ThatIntroduction. . .19the system was more efficient is recognized. That some of the changes were positiveand have opened up room for progressive reform is also within the realm ofdiscussion. What is rejected, however, is the proposition that the new productionmodel, whether we dub it “lean,” “innovation-mediated,” or “post-Fordist”represented a qualitative step forward for labour.The current debate about the new production paradigm is complex and the lackof a rigorous theoretical approach has compounded the problem. In particular, issuessuch as the articulation between forms of work organization and specific features ofindustrial relations, and defining what terms like “Fordism” actually mean, demandclarification.III. History: Critical and Comparative ApproachesTo move beyond the three pillars stereotype, to provide some depth to ourunderstanding of work in Japan, we need to develop a more comprehensiveunderstanding of the postwar history of labour-management relations. Furthermore, itwill be helpful to take advantage of recent theoretical insights into labour process andindustrial relations under capitalism. And, although this study is not, strictlyspeaking, a comparative study of two or three countries, it is still necessary to groundthe discussion with an explicit, comparative reference point in order to avoidinappropriate assumptions.The contemporary themes and debates introduced above have shaped the issuesIntroduction. . .20addressed in this study. But first and foremost this is a historical study of labourrelations as they evolved at the point of production. Earlier research into the coalindustry in Japan acquainted me with the Miike coalminers in Kyushu and theirmomentous fight for survival that galvanized the whole country in 1960. The relativeabsence of strike activity in the post-1960 era, it seemed, was a recent development.I became convinced that no account of contemporary labour-management relationswas worth its salt, if it could not document and analyze the role workers and unionsplayed in the postwar period and if it could not explain the tremendous labour-management conflicts that marked the 1945-1960 period.This, then, is a historical study of industrial relations and productionmanagement in the 1945-1973 period. This particular period was chosen because,despite important continuities between the pre- and postwar period, defeat in W.W. IIand the American occupation marked the beginning of a new era for Japan ineconomic, social and political terms. The cutoff is about 1973-75. By this time,Japan had put its own stamp on its development and many of the contemporaryfeatures of industrial relations and production management were installed.Furthermore, the 1973-1974 oil crisis marked an important socio-economic watershedwhich deserves special treatment that this study could not hope to accomplish.Finally, twenty years seemed a minimum of distance necessary to obtain somehistorical purchase on the often slippery slopes of socio-economic analysis.Introduction.. .21A Crisis of TheoryIt is one thing to use history to challenge current assumptions; it is quiteanother thing, however, to overcome them. I suspect that one of the reasons why thethree pillars typology persists, despite numerous recent qualifications and challenges,is because, put simply, there has not been an adequate theoretical framework toconstruct an alternative, coherent analysis. And so we in the English-speaking worldhave, for the most part, been left to debate Japan’s institutions from the outmoded orlimited perspectives of convergence theories, late development, or strategic choice,among others.32 That a radical framework of analysis has been, until recently, nextto non-existent in the English literature on Japan will hardly come as a surprise. Andeven where it has developed, the radical perspective has not been without itsproblems.Braverman did resuscitate an interest in radical analysis of labour process. Hiswork has achieved classic status and inspired sustained research in the specific area oflabour process.33 Unfortunately the same cannot be said for the arena of industrialrelations. The Marxist tradition has until recently failed to articulate a cohesivetheory that speaks directly to the issue of “industrial relations.” In some cases thishas gone to the point of even questioning the plausibility of industrial relations as aspecific field:32 For an overview of these perspectives see Chapter 1 of Greg Bamber& Russell Lansbury, International and Comparative Industrial Relations,(Sidney, Allen & Unwin, 1987), pp. 3-29.. Harry Braverman, Labour and Monopoly Capital, (New York, MonthlyReview Press, 1974)Introduction.. .22To argue thus would be to accept the theoretical coherence of‘industrial relations’ as an area of analysis: to endorse the materialand theoretical autonomy of institutionalized management-unionrelations. For the same reason, any search for a radicalredefinition of ‘industrial relations’ must be self-defeatingThis narrow Marxist approach has perpetuated a dichotomy between labour processand industrial relations theory. Marxists study labour process or political economy,academics study industrial relations. Fortunately, the abyss is beginning to disappearas capital globalization stimulates international and comparative studies. These, inturn, are forcing Marxism into the twentieth century. This study suggests that onepart of that transition will consist of redefining what is labour process. It is notsimply the study of the division of labour, that is, how work is organized.Regulation theory developed in France, and Burawoy’s theory of productionregimes are two key elements that have helped to resuscitate a constructive Marxistcritique of capitalist development and can help us better comprehend the dynamics ofthe labour process. They constitute the theoretical heart of this thesis. The formerhas provided a Marxist economic analysis that, by challenging traditional Keynesianeconomic theory, permits a deeper understanding of Fordism, that is, the regimes ofR. Hyman, “Theory in Industrial Relations: Towards a MaterialistAnalysis,” in P. Boreham and G. Dow eds., Work and Ineguality Vol. 2:Ideoloqy and Control in the Labour Process, (Melbourne, Macmillan, 1980),p. 55)My interpretation of regulation theory is based on MichelAglietta, A Theory of Capitalist Regulation, (London, Verso, 1987edition); Alain Lipietz: The Enchanted World: Inflation, Credit and theWorld Crisis, (London, Verso, 1985 translation); Towards a New EconomicOrder: Postfordism, Ecoloqy and Democracy, (Cambridge, Polity Press, 1992translation); and‘Towards Global Fordism?” in New Left Review, (No. 132,March-April 1982); and Robert Boyer, The Regulation School, (New York,Columbia University Press, 1990) . On production politics see MichaelBurawoy, The Politics of Production (London, Verso, 1985)Introduction. . .23intensive capital accumulation and the mechanisms that govern them. Burawoy, onthe other hand, posited a theory of “relations in production.” These relations arisenot only from the nature of work organization but also from class conflict andcompromise which gives shape to specific production regimes on a sectoral andnational scale.The combination of these two strains in Marxist theory permits, indeeddemands, a perspective that embraces both convergence and divergence inunderstanding capitalist development on an international scale. They will help usunderstand how industrial relations in postwar Japan facilitated the innovations inproduction process, yet did so without empowering Japan’s workers. To fullyunderstand Japan’s system, however, we must first backtrack and, using the insightsof theory, explore the origins and mechanics of Fordism as it initially developed.Fordism as Labour ProcessThe decline of industrial America and the surge of Japanese investment in theform of automobile transplants erected in North America have heightened interest inthe study of production management and work organization. The pierre de touche inthis discussion has been the production system that Henry Ford introduced into hisoperations with the Model T in 1908-1913. These changes culminated with theintroduction of the chassis assembly line production at Ford’s Highland Park plant inIntroduction. . .241914.36 Most commentators today agree that the Ford system was, in its embryonicform, the prototype of modern mass production. It was based upon “deskilling, butalso product standardization, the use of interchangeable parts, mechanization, amoving assembly line, and high wages.”37 This latter quote, in fact, represents whatwe might call the popular or functional definition of Fordism.What had happened in the automobile industry, in fact, was a revolution in theorganization of work based on the capitalist imperatives for ever-increasing efficiencyand a technical revolution in steel manufacturing. The development of high gradesteel and Ford’s insistence on using standard gauges created the possibility for thestandardization and interchangeability of parts. Ford’s managers used thesedevelopments to push the division of labour. According to Ford, time study of thework process revealed, for example, that assembly of pistons and rods for engineswith one worker doing the whole process required nine hours of which fours hourswere consumed in walking to fetch or move parts. The work was reorganized sothat: “Instead of one man performing the whole operation, one man then performedonly one-third of the operation--he performed only as much as he could do withoutshifting his feet.”38 In 1908, just prior to the introduction of the Model T, theaverage fitter’s cycle time--the time between repetition of the same operation--was36 The most detailed study of this process is Stephen Meyer III, I1IFive Dollar Day: Labor Management and Social Control in the Ford MotorCompany, 1908-1921, (Albany, State University of New York Press, 1981)Ruth Milkman, “Labor and Management in Uncertain Times,” in A.Wolfe, ed., America at Century’s End, (1990), p. 134. This citationcontains what I term the functional description of the generic Fordistregime.38 As cited in Stephen Meyer III, The Five Dollar Day, p. 21.Introduction. . .25514 minutes.39 Under the new system, however, Ford re-organized the productionprocess so that workers narrowed their job content into short, repetitive operationsuntil the cycle time was reduced to 2.3 minutes in August, 1913 just prior to theintroduction of the assembly line.40The nominal productivity increases were impressive and the new division oflabour required Ford to hire thousands of new workers. Realizing the potentialproductivity gains was not so easy. Co-ordinating production to capture theeconomies of scale was not easy. Furthermore, the intense pace and routinization ofwork created new bottlenecks as automobile workers demonstrated their resistancewith their feet. Absenteeism and turnover at Ford reached astronomic levels. Toaddress this issue, the automaker introduced in 1914 the five dollar day, a profitsharing/bonus scheme that effectively doubled wages for semi-skilled workers. Theonly hitch was that Ford workers had to agree to submit to an investigation by Ford’ssociological department. This group conducted home visits to ascertain the moralcharacter of employees. This paternalistic regime only lasted until 1920 at whichtime the sociological department was dissolved and Ford instead combined the carrot,a year-end bonus system based on skill and length of service, with the proverbialstick--a network of spies to report on slackers and union organizers.41 A pointseldom emphasized, however, is that Ford’s wage system had for the most part. James P. Womack, Daniel T. Jones & Daniel Roos, The Machine thatChanged the World, (New York, Macmillan Publishing, 1990), p. 28.° James P. Womack, Daniel T. Jones & Daniel Roos, The Machine thatChanged the World, p. 28.‘ Ibid., p. 197.Introduction.. .26abolished any form of individual or collective production bonus or incentive paymentthat was tied to overall output. Instead wages were paid on a straight time basis withsome adjustment for skill that was incorporated into the wage scale through aclassification system tied to specific jobs.42Ford’s production system was a prototype, and short cycle times, repetitive jobroutines and detailed operations charts dictated by management and the industrialengineer became legion in the U.S. automobile industry. Little has changed since.Meyer, for example, documented the changes in the Fordist regime within the U.S.auto industry after 192O. The two major changes he points to were theintroduction of flexible specialization allowing for annual model changes underG.M. ‘s president, Alfred Sloan, as early as the 1920s, and the advent of extensiveautomation in the postwar period prompted by union institutionalization of high wagesand influence over job assignments. Despite these innovations, Meyers concludedthat “Fordism remained a managerial strategy for the control of workers and thereduction of labour costs.” For workers, this resulted in a “diluted skills, intensifiedwork, and eliminated possible jobs. “The Fordist norm for work organization, the type of repetitive and routinizedwork that developed in the automobile industry was a classic example of what became42 For details of the new job hierarchy see Stephen Meyer III, IiFive Dollar Day, pp. 101-104.‘. Stephen Meyer, “The Persistence of Fordism: Workers and Technologyin the American Automobile Industry, 1900-1960” in Nelson Lichtenstein &Stephen Meyer eds., On the Line, Essays in the History of Auto Work,(Urbana, University of Illinois Press, 1989), pp. 73-99.Stephen Meyer, “Persistence of FordismlT, p. 94.Introduction. . .27known as “Taylorism.” At the same time as Ford was introducing the new form ofwork organization at Highland Park, others were conducting time and motion studiesof the work process in sectors outside the automobile industry. The most famous ofthese consultants was Frederick Taylor.45Frederick Taylor and his associates Henry Gantt, Carl Barth and Horace Hathaway were active in the American Society of Mechanical Engineers at the turn of thecentury. They began to do contract work for employers, scrutinizing the workprocess of both craft workers and labourers and submitting them to rigorous standardsof efficiency.46 The movement grew in scope and in 1911 Taylor and his manyassociates founded the Society to Promote the Science of Management (renamed theTaylor Society in 1915 upon Taylor’s death). The growth of Taylorism and‘scientific management’ signified the elevation of the study of the labour process to aseparate discipline in order for employers to gain complete control over how workwas organized and to maximize efficiency through exact instruction of detail work.To be sure, as Braverman indicated in his seminal study of labour process, thelong term impact was to reduce the control and power of the craft worker. Thecontinuous redivision of labour in the early 20th century had tremendous reperTaylor himself acknowledged that Ford developed the repetitiveand restricted work procedures independently of Taylor when hecongratulated 600 Detroit automobile industry managers for being the‘first to install the principles of scientific management without the aidof experts.” Stephen Meyer III, The Five Dollar Day: Labor Management andSocial Control in the Ford Motor Company, 1908-1921, p. 20.. For a mainstream account of Taylor’s experiments see DanielNelson, Frederick W. Taylor and the Rise of Scientific Management(Madison, University of Wisconsin Press, 1980)Introduction. . .28cussions on the composition of work skills. Non craft workers began to predominatein assembly operations and, as David Montgomery points out: “Skilled workers inlarge enterprises did not disappear, but most of them ceased to be productionworkers. Their tasks became ancillary--setup, troubleshooting, toolmaking, modelmaking---while the actual production was increasingly carried out by specializedoperatives.”47 Clearly, Taylorism and scientific movement had a profound impacton the organization of work, particularly in mass production industries. But for anumber of reasons its influence remained partial.For one thing, Taylorism provoked organized resistance from many unionworkers. Moulders at the Watertown Arsenals walked out in protest of Taylor’sexperiments and prompted a congressional investigation of Taylorism. Congress infact prohibited time studies in government arsenals and navy yards in 1915!48 Thescale and nature of the production process varied according to product and size ofmill; this left some room for the skilled worker and restricted the ability of managersto exert control through work organization. Productivity growth called forth hugeincreases in the number of workers in certain crafts. Furthermore, the redivis ion oflabour also created new forms of crafts and skills over which management had onlypartial control.49 And as movements, Taylorism and scientific management‘. David Montgomery, The Fall of the House of Labour, (Cambridge,Cambridge University Press, 1987), p. 215.. Ibid., p. 221.. In Labor and Monopoly Capital, Braverman posits that these trends“simply mask the secular trend toward the incessant lowering of theworking class as a whole below its previous conditions of skill andlabor.” (pp. 129-130). This would seem to be overstating the case andignores the dynamics of capitalist productivity. It also sets linearIntroduction. . .29themselves went through important transformations, particularly after Taylor’sdeath.5° Despite these limits, many aspects of scientific management persistedmainly in the modern guise of industrial engineering. Its symbols were, and continueto be, the stop watch, time-and-motion studies, and the detailed operations chartwhich remained prevalent especially in the automobile industry.The relevance of this discussion to the work at hand is twofold. First, anumber of Japanese industrial engineers such as Ishikawa Kaoru contend that Japaneseemployers did not adopt Taylorist work methods. As well, a number of Westernscholars, such as Kenney and Florida, assert that the Toyota production system (leanproduction) has broken with the Fordist labour process and reached a new level ofpost-Fordist development. The evidence from this historical review of labour processin the automobile industry (summarized in Chapter 6) indicates that, up to 1975 atleast, Taylorist forms of work organization were alive and well in the automobileindustry, as was assembly line production. We discovered, however, significantdifferences between Fordism as it evolved in Japan and in the United States but nonethat would lead us to qualify the Toyota system as post-Fordist. The variations,including flexible production and extensive employee involvement through qualitycircles, were significant and can be attributed to a number of factors. Early postwarstandards of skill (handicraft equals best, operative equals worst) whichcan easily be interpreted to mean that the goal of even modern labour iscraft production methods.For details see chapter 10 in David Noble, America by Design,(Oxford, Oxford University Press, 1979) and Steve Fraser, “The LabourQuestion,” in Gerstie & Fraser eds., The Rise and Fall of the New DealOrder, 1930-1980, (Princeton, Princeton University Press, 1989), pp. 55-84.Introduction. . .30circumstances in Japan did not allow for the direct application of mass productiontechniques as they were being used in U.S. automobile plants, for example. Smallbatches had to be integrated into a continuous flow process. Furthermore, whileemployers were enamoured with the mass production system and Taylorism, they didnot necessarily embrace the industrial relations practices (the wage system, jobdescriptions, and so forth) that had evolved under the influence of the U.S. unionmovement. They were therefore able to use labour more flexibly. As these examplespartially illustrate, understanding the distinction between work organization andindustrial relations and, at the same time, their inter-relationship is crucial forunderstanding convergence and diversity in Fordism as it evolved in Japan and otherparts of the world.Fordism and Regulation TheoryFrench Marxists Michel Aglietta and Alain Lipietz went beyond a functionaldescription of Fordism to articulate a general theory of capitalist regulation based on ahistorical assessment of the U.S. experience. They made the following points:1) Fordism embraced and went beyond Taylorism through the use of the semiautomatic assembly line which became a core component of the new labour process.It created a new benchmark for continuous flow operations that required, in theautomobile industry at least, the standardized, repetitive cycle of movements overwhich labour had almost no control;Introduction. . .312) the resulting productivity increases articulated a new relationship betweenprocess of production and mode of consumption. In other words, Fordism bothcreated and demanded the development of mass consumption but employers’ short-term perspective (anti-labour bias) prevented the establishment of the massconsumption norm until after the depression;3) the articulation of an independent labour agenda through unions andpolitical parties demanded a new form of relationship between capital and labour.This new relationship was hegemonic, that is, based on the consent of labour to thecontinuing existence of the regime in return for an independent voice at theworkplace.4) the necessity of mass consumption and stable labour-capitalist relations gaverise to the Fordist state that, to one degree or another, regulated industrialdevelopment, socialized a part of the expense of reproducing labour power (throughsocial insurance, schooling, health care and so forth), and created a framework forhegemonic labour-capital relations.5) in economic terms, mass consumption created a balance between theproducer goods and consumer goods sections of the economy (heavy and lightindustry) and facilitated the passage of capitalist accumulation from an extensive phase(extension of working hours) to an intensive one (accelerated labour processdependent on ever-increasing investment in fixed assets);6) problems in the labour process remained, including balancing the assemblyline (standardizing each work routine to a specified period), negative effects ofIntroduction.. .32routinized work and speed-ups on the labour force, and the potential dangers ofcreating a workforce with a shared experience.Figure 1.2: Regulation Theory and FordismFordist labour process:a)parts and products are standardized;b)job routines are broken down into routinized cyclesc)the assembly line institutionalizes and extends the Fordist labour process;d)management retains control over the labour process although control may be restricted by thenature of labour-management relations;e)automation is perceived as the main means of improving productivity and reducing labourcosts.Fordism’s social dimensions:f)mass consumption as a requisite norm;g)socialization to some degree of the reproduction of labour power;h)a social contract of some sort between organized labour and management that embraces theessentials of Fordism.Fordist economic dynamics:i)a relative balance between the producer and consumer sectors of the economy;j)a general if segmented increase in wages that is maintained through regular salary increasesfor organized labour.Regulation theory offers significant advantages as a labour-based, inclusive andcomprehensive framework for understanding capitalist development. It affords theantagonistic wage relationship between labour and capital a centrality which is deniedby conventional social sciences and is at times ignored by some state-orientedMarxists. By describing it as inclusive, I mean it acknowledges that labour,employers and the state were all key players in elaborating the specific regulatorymechanisms necessary for capital accumulation. For example, it correctly identifiesthe historical role of labour’s struggle for a shorter work week as a key determinantin pushing employers to find alternative means of accumulation through extension ofIntroduction. . .33the division of labour and mechanization.Regulation theory also illuminates the complex nature of capital accumulation.The locus of regulation can be found not only in the labour process or in the state butalso in the modalities that link production and consumption, in the ways in whichlabour and capital interact at various stages of the valorization of capital. In otherwords, it allows for the integration of politics and economics at every level.Regulation theory also adopts the Gramscian notion of hegemony, that is thatcapitalist control can no longer be exercised by authoritarian means alone and thatsome sort of deal must be worked out between labour and capital. In other words,wage workers, through their union or shop representatives, must accept many facetsof an oppressive industrial order in return for some say over the terms of employmentand working conditions. Labour’s concessions and influence are often contained inwork rules or collective agreements. Thus, while struggling against exploitation,labour also consents to it, but this consent is derived from a coercive economicsystem based on private ownership of the means of production. As we shall see,Japan also passed into the Fordist phase of capital accumulation but this did not occuruntil the 1960s.Although not always conducted under the rubric of regulation theory, the studyof national variations in Fordist regulatory mechanisms has become the focus of anincreasing number of studies. Charlotte Yates and Nelson Lichtenstein have helpedestablish one important point of demarcation in Fordist regimes through theirIntroduction. . .34respective studies of labour in Canada and the United States.5’ Their workilluminates how Canada and the United States were unable to pursue the social-democratic route that culminated in the corporatist mediation (labour/business/state)that became the hallmark of Fordism in Sweden, for example. Yates terms this noncorporatist model a “liberal-pluralist” form of regulation in which the regulatorymechanisms are integrated into and diluted by the private market and notuniversalized by public control through the state. Such insights are suggestive andcan help to explain the particular form of Japan’s Fordism. They are related to theattempts by Michael Burawoy to develop a theory of production regimes.Burawoy and the Politics of ProductionRegulation theory has directed our attention to the universal aspects ofFordism as an intensive regime of accumulation which demands the creation of theregulatory state. Burawoy on the other hand directs us to examine the variations inthe nature of production regimes as they are reproduced at the workplace.Burawoy begins his thesis by refuting Braverman’s proposal that the fundamental aspect of capitalism is its control of the labour process through the division oflabour and the concomitant division of conception from execution. While upholdingthe classic status of Braverman’s work as critique, he contends that it cannot stand asa framework for analysis. Braverman misses the essence of capitalist control because‘. Charlotte Yates, From Plant to Politics: The Autoworkers Union inPostwar Canada (Philadelphia, Temple University Press, 1993) ; NelsonLichtenstein, “From Corporatism to Collective Bargaining,” in Fraser andGerstle eds., The Rise and Fall of the New Deal Order, 1930-1980(Princeton, Princeton University Press, 1989)Introduction.. .35his framework remains within capitalism: “By contrast, Braverman takes hisstandpoint from within capitalism, alongside the craft worker--the embodiment of theunity of conception and execution.”52 Braverman also fails to capture a relativenotion of capitalist control because he does not articulate a potential alternative modelof worker control. Socialism “is deduced for Braverman by inverting a picture ofcapitalism taken from within.”53 In contrast to Braverman, Burawoy contends:Capitalism can and did survive under conditions of the unificationof conception and execution. Their separation is not at the core ofthe capitalist labour process per se but is something that emergesand disappears in an uneven fashion as capitalism develops. Thecraft worker was, and indeed in some places still is, a part ofcapitalism. Thus, to identify the reunification of conception andexecution with socialism is to confuse job control with workerscontrol, relations in production with relations of production. Itrisks not going far enough and, in the process, mistaking anostalgia for the past for a nostalgia for the future.54Burawoy demands that we go beyond Braverman and broaden our understanding oflabour process. First we must stop reducing the labour process to the simple questionof the division of labour and work organization.55 The labour process encompassesboth the organization of work and what Burawoy terms the ‘relations in production,’that is, the apparatus of production which regulates labour-management relations atthe point of production. Thus any examination of the labour process must entail not52• Michael Burawoy, The Politics of Production, p. 23.Ibid., p. 24.Ibid., p. 54.. Burawoy is at times confusing on this issue, sometimes equatingwork organization and labour process himself--see The Politics ofProduction, p. 8 versus the definition on page 31.Introduction. . .36only how work is organized but the entire scope of employment relations includingwage and job determination, hours of work, dispute resolution and so forth.56Another critical factor is correctly discerning the relationship between labourprocess and specific factors which might condition it. For example, labour processcan be affected by the nature production in specific industries, inter-firm competition,technological developments, labour market conditions, gender and race issues,ideology, not to mention the degree and nature of state intervention.Taking a broader approach to labour process, and allowing for a multi-factoranalysis of variables that might impact on it, allows us to open up and examine thevariations in capitalist development over time and space. Burawoy does this for bothold capitalism--the regimes of the early industrial period--and for advanced capitalism,the hegemonic regimes.In his analysis of the former, Burawoy concludes that Marx was incorrectwhen he implied that capitalism could only give rise to one type of regime--marketdespotism. Many types of production regimes existed even in the early period ofcapitalism including the ‘company state’ (early throstle mills in England); paternalism(Lowell Mills 1830-1860); patriarchy (mule spinning in England); and marketdespotism (New England mills after 186O). More relevant to the discussion athand is Burawoy’s characterization of advanced capitalist regimes.. Burawoy saw this through his comparison of work at Allied andJay’s. Even though the work process, that is what people did in themachine shops, was organized in a similar manner, the relations inproduction were quite different.. Michael Burawoy, The Politics of Production, p. 91.Introduction. . .37Burawoy places considerable emphasis on the role of the state in shaping theregimes of production in advanced capitalist countries. Using the development ofstate welfarism (support for reproduction of labour power, that is, maintenancesupport that allows workers to live even without employment) and the degree of directstate regulation in production as variables he comes up with a schematization asillustrated in Figure 1.3. This comparative framework is based exclusively on thedegree of state intervention and is not, therefore, a comparison of regimes per sewhich must take into account other factors. Nevertheless, Burawoy’s schematizationis interesting in that he poses Japan and Sweden as twoFigure 1.3: State Intervention and Factory RegimesDegree of State Support forthe Reproduction of LabourPowerHigh LowHigh Sweden U.S.Degree of Direct StateRegulation of Factory Regime Low England JapanSource: Michael Burawoy, The Politics of Production, (London, New LeftBooks, 1985), p. 138.opposite poles in relation to state intervention in the labour process. The findingsfrom this study confirm Burawoy’s perspective and refute’s Dore’s association ofJapan with Sweden. In trumpeting about the benevolence of Japan’s employers andthe progressive aspects of Confucianism, Dore desperately attempts to find a bridgingmechanism when in fact the two regimes are quite different!The last points concerning the theory of variable production regimes areIntroduction. . .38related to concepts of the state and notions of employer control. The general thrust ofBurawoy’s analysis is to combat what he considers to be the under-politicization ofproduction and the over-politicization of the state, that is, theories that stress thestate’s “autonomy, dislocating it from its economic foundations.”58 Burawoy sees anorganic link between production apparatuses and the state and even goes so far as tosuggest that, from a historical perspective at least, the former determine the shape androle of the state. This study confirms this theory and goes a little further. Japan’spolitics of production definitely shaped the Fordist regime of the 1960s. The keyquestion from labour’s perspective, however, was (and is) to what degree did theproduction regime, or the state, allow the separation of the reproduction of labourpower from the market and employer control?59 In simpler terms, to what degreecan labour de-commodify itself within the confines of a commodity-based capitalistsystem. This, I contend, is the index against which we can assess the role of the stateand also the particular nature of any given production regime. In Japan, as we shallsee, the divisions in the labour movement and the ascent of enterprise unionismweakened labour’s ability to de-commodify itself, both at the level of the state and theworkplace.Notions of capitalist control can neither be reduced to a single dimension norunderstood statically. Capitalism can only survive so long as capital is capable ofextracting surplus value. For this it needs labour. But because labour resists. M. Burawoy, The Politics of Production, p. 122.Gosta Esping-Jndersen designates this separation as the decommodification of labour. See his work, The Three Worlds of WelfareCapitalism, (Princeton, Princeton University Press, 1990)Introduction. . .39exploitation, capital would prefer to rid itself of labour or control it absolutely, bothof which are impossible. As a result we have the ultimate in love-hate relationships.There are thus both economic and political aspects to the labour process and theessence of capitalist control is, as Burawoy puts it, to secure surplus value while atthe same time keeping it hidden. This is an important formulation because it capturesboth the economic and political moment of capitalist relations. Employers mustsecure the maximum surplus value possible but at the same time not overly expose thehidden exploitative relationship inherent in the enterprise. This tension was indeed atthe heart of production politics in Japan and, using these theoretical insights, thisstudy traces the evolution of the relations in production in Japan (industrial relations)and attempts to articulate the impact of the dominant pattern on the organization ofwork, particularly in the automobile industry.IV. A Specific Comparative FrameworkBefore proceeding to the body of research, I propose to offer a specific spatialreference for comparative purposes. This is necessary for two reasons. First, Fordisttheory as we have discussed it has been celated mainly to the issue of workorganization or to a general theory about capital accumulation. Second, the impact ofspecific forms of industrial relations can only be fully understood when they areevaluated from a specific, comparative perspective. We need, therefore, a referencepoint in industrial relations. Indeed, understanding different regimes often requiresthe destruction not of a single stereotype but a symbiotic set, the thesis and anti-thesisIntroduction. . .40often associated with different countries. Take for example unions: To many, the unitof Japanese trade unions is the ‘enterprise union’ while its counterpart is described asan ‘industrial union’ as supposedly epitomized by national unions in the United States,such as the United Auto Workers (UAW). The juxta-positioning of the two helps tocreate a competitive dynamic, to attempt to define one or the other as superior. Buthow quickly we can slide into assumption! Someone counterposes the termsenterprise and industrial union and presto--we have supposedly defined them both! Apriori deduction at best, the problem is amplified when we juxtapose two entities andassociate them with two different countries. The reality is much more complex andboth enterprise and industrial unions come in a variety of configurations that defysimplistic comparisons. Thus even before we delve into the history of Japan’sproduction regimes it seems essential to provide a modicum of detail about ourcomparative frame of reference.Given the limitations of this work, I have chosen Canada and the United Statesas the specific points of reference. The reasons for this are fairly obvious and have, Ishould state, nothing to do with convergence theories. Neither the United States norCanada is the gold standard! But they will serve as a reference point for a number ofreasons. First, it was in these two countries that Fordism as a system first matured.Second, they are the regimes which I know best and the reader and I can avoidinherent assumptions if I state my own perceptions about these regimes. Finally, thechoice of the United States seems appropriate given the extraordinary influenceAmericans attempted to exert on Japan through the Occupation period and the closeIntroduction. . .41economic ties that developed thereafter. In discussing these two countries’ regimes,however, I shall attempt to provide some further international comparisons as well.Historical BackgroundKim Moody has argued that the major features of labour-capital relations inthe U.S. since W.W. II have been “national pattern bargaining, grievances proceduresdesigned to remove conflict from the shop floor, and bureaucratic unionism.”60 Ashistorical polemic, Moody’s arguments are powerful and extremely useful in explaining some of the weaknesses that have dogged the U.S. labour movement in the pastdecade. As comparative frame of reference, however, Moody’s analysis requiresmodification.61 For the purposes of this study, I contend that the primary features ofthe industrial relations system in the United States and, to a large extent, in Canada60 Kim Moody, An In-lury to All, the Decline of American Unionism,(London, Verso, 1988), P. 20.‘ This account of labour relations is based on the following works.For the United States: David Brody, Workers in Industrial America (NewYork, Oxford University Books, 1993 edition); C. Gersuny & G. Kaufman,“Seniority and the Moral Economy of U.S. Automobile Workers, 1934-1946,”Journal of Social History, (Spring, 1985), pp. 463-475; Harry C. Katz,Shifting Gears (Cambridge, MIT Press, 1987); Edward Levinson, Labor on theMarch, (New York, University Books, 1956 edition); Fraser & Gerstle eds.,The Rise and Fall of the New Deal Order, 1930-1980, (Princeton, PrincetonUniversity Press, 1989) ; Nelson Lichtenstein and Stephen Meyer eds, On theLine, Essays in the History of Auto Work (Urbana, University of IllinoisPress, 1989); Leon Litwack, The American Labour Movement (New York, Simon& Schuster, 1962); David Montgomery, The Fall of the House of Labour,(Cambridge, Cambridge University Press, 1987); Kim Moody, An Iniury toAll, the Decline of American Unionism, (London, Verso, 1988) . For Canada:John C. Anderson et al, Union-Management Relations in Canada, (Don Mills,Addison-Wesley, 1989); L.S. MacDowell & I. Radforth, Canadian WorkingClass History: Selected Readings, (Toronto, Canadian Scholars Press,1992); Desmond Morton, Working People, (Toronto, Summerhill, 1990edition); Bryan D. Palmer, Working-Class Experience, (Toronto, Butterworth, 1983); Gerald S. Phillips, Labour Relations and the CollectiveBargaining Cycle, (Toronto, Butterworths, 1981); James Rinehart, Thefyranny of Work, (Toronto, Harcourt Brace Jovanovich, 1987 edition)Charlotte Yates, From Plant to Politics, (Philadelphia, Temple UniversityPress, 1993)Introduction. . .42include:> the partial triumph of industrial, as opposed to craft unions, thatallowed for the organization of all workers on the basis of ‘one shop-one union;’> a regime of compulsory union recognition in exchange for abureaucratic/legalistic method of dispute resolution;> institutionalized collective bargaining on wages and working conditions every one to five years in which single enterprise bargaining is the norm butcentralized bargaining in the form of pattern or joint-bargaining also occurs;> extensive and legalistic collective agreements that have grown upbased on a system of industrial jurisprudence in which residual managerial rightspredominate (i.e. management controls anything not spelled out in the collectiveagreement);> an occupation/classification-based wage system that operated on theprinciple of “equal pay for equal work” and comparability and under which incentivesystems play a secondary role;> union job controls (“restrictive work practices”) including extensiveseniority rights, detailed occupational classifications, job descriptions, bumping rights,and so forth, enforceable by shop stewards but subject to bureaucratic grievanceprocedures in cases of dispute.This system of contemporary industrial relations in Canada and the UnitedStates congealed in the 1935-50 period, although there were important supplementarydevelopments in the 1960s. The Fordist mechanisms of workplace regulation thatIntroduction.. .43édeveloped in the two countries were substantively similar with some notableexceptions. It is difficult to do justice to years of work history in a few pages, evenfor one country let alone two! But for comparative purposes some of the key pointsare highlighted in the following pages.hidustrial Relations: U.S. and CanadaHenry Ford had revolutionized the labour process with the changes broughtabout by the re-division of labour and the institutionalization of the changes throughthe introduction of the assembly line. This restricted the development of craft skillsand obliterated individual control over work. At the same time the new form of workorganization substantially increased productivity and created the conditions necessaryfor the incorporation of millions of workers into production. These changes initiallyoccurred in the 1908-1920 period, yet Fordism did not mature at this time. On thewhole, employers, including Ford himself, remained dedicated to the open shop andmaintained a fundamentally antagonistic attitude towards collective bargaining.62High wages never became institutionalized and this, among other factors, precipitatedthe economic crisis that began with the crash of 1929.The depression brought forth the Roosevelt administration in the United Statesand it passed the National Recovery Act in 1932 to be followed later by the WagnerAct in 1935. The reforms represented by these two acts constituted a watershed inU.S. regulation of labour relations. They marked the triumph of a new vision of62 See David Montgomery, The Fall of the House of Labor, pp. 269-275.Introduction. . .44industrial relations which, although not supported by a majority of employers by anymeans, became the standard for decades to come. This new standard provided forcompulsory employer recognition of independent unions on the condition they fulfilledcertification requirements.Historical circumstance played an important role in establishing the newregime. Mass production industries had created a new type of working class whichtraditional craft unionism, as typified by the A.F.L., was unwilling to embrace. Theeconomic backdrop to this legislation was of course the depression and the subsequentpolitical perception that capitalism, left unregulated, was unable to sustain itself. Thisput employers on the defensive and created the momentum for the election ofRoosevelt. But the labour relations component of the New Deal had been formingeven prior to Roosevelt’s election. As Steve Fraser has shown, even in the 1920s asmall minority of employers and consultants (many of them in the Taylor society!)had begun to articulate a new mode of regulation which accorded independentrepresentation for workers through unions and collective bargaining.63 These peopleworked with union leaders such as Sidney Hiliman of the mens clothing union inelaborating a new labour-management deal. The essence of that deal was anacceptance of scientific management and employer rights tempered by a bureaucraticform of regulation encompassed by a collective agreement.Capitalism’s first failure, the Great Depression, precipitated the meteoric rise63 Steve Fraser, TiThe ‘Labor Question,TT in S. Fraser & G. Gerstieeds., The Rise and Fall of the New Deal Order, 1930-1980, (Princeton,Princeton University Press, 1989), pp. 55-84.Introduction. . .45of regulationists within Roosevelt’s New Deal administration. But as historians suchas Lichtenstein and Brody have documented, it remained tough slogging for the labourmovement even after passage of the Wagner Act in 1935 and the establishment of thedO in 1936-37. Only U.S. entry in W.W. II created the exceptional conditions thatallowed the state to actively promote the consolidation of unions and regulate wageand prices. This was the pinnacle of the ‘corporatist’ wedge within the liberal U.S.state.The 1946 defeat of Walter Reuther’ s autoworkers bid for a wage increasewithout an increase in the price of cars, and the enactment of the regressive Taft-Hartley Act in 1947 marked the end of the wartime regime and the resurgence ofconservative business in the postwar era. The failure of the ClO to break with theDemocratic Party in 1947-1948 marked the incorporation of progressive unionism intoa postwar order based on “alignment with the government in the battalions of the newcold war and exclusion of the Communists from the political arena.In Canada’s case, legislation similar to the Wagner Act was only introduced in1944. McKenzie King, leader of the Liberal government, had been long associatedwith the company union movement and was adamantly opposed to compulsoryrecognition of unions. It took the Wagner Act in the United States, a 1943 strikemovement that surpassed all previous levels of strike activity, and the rise of thesocial-democratic Co-operative Commonwealth Federation (C. C. F.) to convince Kingthat it was time to embrace the politics of the New Deal, including union recognition.“. Nelson Lichtenstein, TThe Eclipse of Social Democracy,” p. 141.Introduction. . .46Privy Council order 1003 (PC 1003), that more or less embodied the Wagner Act,was enacted in April 1944.The essence of both pieces of legislation was that employers were,theoretically at least, obliged to recognize and bargain with unions which obtainedcertification as worker representatives. In return, unions were obliged to agree toinclude in their collective agreements provisions for dispute resolution which wouldprevent job action during the life of the agreement.65 Prior to this, unions existed ina state of limbo. Although many of the master-servant or conspiracy laws that hadbe-devilled unions in the 19th century had been struck down, employers were notobliged to recognize or bargain with unions. The Wagner Act and PC 1003 changedthe political climate, but the labour movements in both Canada and the U.S. had towage relentless struggles to make use of the new legislation.Mass production, depression, new labour legislation and war converged tocreate the context for the partial triumph of industrial unionism that took place in the1937-50 period. The first upsurge in the organization of the mass production indus. Two important points of divergence between the U.S. and Canadianexamples must be noted. First, in the Canadian instance, provincesexercised almost exclusive jurisdiction over labour relations after 1925.PC 1003 applied to most private sector workers only because the federalgovernment had appropriated much provincial power through federal wartimecontrols. However, most provincial governments in Canada passedlegislation similar to the Wagner act after 1948. They did notnecessarily emulate the Taft-Hartley amendments to the Wagner act whichoccurred in the United States and which weakened many of the provisions ofthe Wagner Act. This has resulted in a second divergence in the labour-management environment in Canada, namely somewhat easier certificationprocedures in some Canadian provinces, extensive use ofconciliation/mediation prior to strike action, stricter controls ongrievance procedures, and relatively stronger union security regulations.For details see Donald Carter, “Collective Bargaining Legislation inCanada,” in John Anderson et al, eds, Union Management Relations inCanada, pp. 34-35.Introduction. . .47tries began in the 1935-37 period in both the United States and Canada. This earlyspurt was more dramatic in the United States but was not sustained in either country.W.W. II was a crucial period for union expansion in the United States and to asomewhat lesser degree in Canada. It was only in the early postwar period (1945-48)that Canadian union densities reached the 30 percent plus figures that had alreadybeen achieved in the U.S. Union densities reached their peak levels (34-35 percent)around 1955 and then declined over the next decade in both countries. Around 1965,union densities in Canada began to recover and hovered close to the 40 percent levelsince 1978 whereas in the United States union densities continued to fall.66The rise of industrial unionism was predicated on a significant shift in unionphilosophy. Prior to 1935, the predominant form of union had been the craft type,that is, organizations of workers with a defined trade. The shift to organize the non-craft workers in the mass production industries required a new organization, theCongress of Industrial Organizations, because of the hidebound haughtiness of thecraft unions (organized in the American Federation of Labour) towards non-craftworkers. The rise of industrial unionism was in opposition to craft unionism. It isquite true that many of the unions, including the autoworkers, steelworkers,As mentioned above, part of the reason for Canadian union vigourcan be attributed to a relatively favourable legal context. This appearsto have had some bearing on the rise of unionism in the public sector inCanada, which most studies point to as one of the most significantdivergence with the U.S. and the reason for sustained union power in thepost 1965 period. A second notable feature of the Canadian situation hasbeen the gradual Canadianization of unions in Canada. Most unions inCanada were so-called !IinternationalTT unions, that is, U.S. unions withCanadian locals. The reign of TTinternational unionsTT in Canada lasteduntil 1977 by which time the majority of union members belonged toCanadian unions.Introduction. . .48rubberworkers, and so forth, were national in scope but then too, many of the craftunions were also national in scope. The essential difference between craft andindustrial unionism was the idea that all workers, regardless of craft, should beorganized in one union whether it be by industry, enterprise or plant by plant.67This dynamic is crucial to any understanding of industrial unions in North Americaand, as we shall see, begins to undercut the organizational dichotomy betweenindustrial and enterprise unions that is so commonly assumed by adherents to thethree pillars interpretation of Japan’s industrial relations. This dichotomy is furtherchallenged if we move to the level of collective bargaining.The unit of bargaining certification issued by labour relations boards in Canadaor the United States was usually not a whole industry but rather the plant or enterprise(single company with multiple plants). Nor should industrial unionism be equatedwith industry-wide bargaining because, with very few exceptions, seldom did industry-wide bargaining ever exist. As a matter of fact, in a Canadian study of bargainingstructures in units of 500 employees or more, less than 20 percent of the cases (40percent of workers) involved multi-employer bargaining in 1965.68 If units of lessthan 500 employees were included the ratio of multi-employer bargaining droppedeven further. In other words, for better or worse, the plant or enterprise remained‘. To understand the organizational principle behind the ascent ofindustrial unionism, see the Minority Report of the Resolutions Committeeon Organization Policies: A.F. of L. Convention (1935) as cited in LeonLitwack, The American Labor Movement, pp. 49-51.. John C. Anderson, “The Structure of Collective Bargaining” in JohnC. Anderson, Union Management Relations in Canada, p. 218. By 1982, thelevel of centralization had dropped further below 1965 levels.Introduction. . .49the centre of industrial relations in Canada and the United States in the postwarperiod.Recognizing the centrality of the plant or enterprise in Canadian/U.S.industrial relations should not blind us, however, to the fact that other forms ofbargaining did develop, and although they did not become the standard, they wereimportant. Except for the construction industry there was almost no legal provisionfor certification of industry-wide bargaining units for either employers or employeesin Canada. In one of the most centralized bargaining units, the coastal lumber millsof British Columbia, employers formed a voluntary bargaining council which signed amaster agreement with the corresponding union bargaining agents. Even thisagreement was supplemented by locally negotiated agreements on plant-level issues.In Canada and the U.S., a well known form of connective bargaining was “patternbargaining” that evolved in the automobile industry. This practice consisted ofchoosing one enterprise, Ford, General Motors or Chysler, as the bargaining targetand subsequently pursuing collective bargaining with the target enterprise, up to andincluding strike action, until a settlement was reached. Similar settlements wouldthen be demanded of the other two major automakers. Pattern bargaining in this casewas a form of multi-plant, enterprise level bargaining. Even under patternbargaining, the agreements struck on the enterprise level (master agreementsconcerning wages, pensions, and so forth) were supplemented by local agreementsregarding working conditions negotiated at the plant level.A third type of bargaining structure involved co-ordinated bargaining. ThisIntroduction. . .50type of bargaining did not culminate in a master agreement. Instead the agreementsreached were incorporated in local agreements. It should be stressed, however, thatthese trends toward centralized bargaining never became the standard and the plantand enterprise remained the organizational centre of collective bargaining andindustrial relations.The outcome of collective bargaining in Canada and the U.S. was the detailedcollective agreement which could run into hundreds of pages. This type of collectiveagreement only developed with the rise of industrial unionism under the legalisticWagner-type industrial relations system. Many of the early craft contracts were onlya few pages long. The detailed collective agreement arose first as a response toresidual rights theory, that is, the theory that what was not in the contract remainedthe prerogative of management and second, in response to the legalistic arbitrationprocess for resolving grievances arising from differing interpretations of the collectiveagreement.The detailed collective agreement contains stipulations regarding hundreds ofitems. Two of specific importance for our purposes are the detailed wage schedulesand job control rules. While many variations emerged according to industry andunion, on the whole wages were pegged to occupations or job classifications.69Incremental steps or a wage ladder often existed, but on the whole unions demandedequal pay for equal work and an end to favouritism in wages. The role of. The job-based wage system, that subsequently gave rise to theoften complex classification codes, dates back at least to Ford’s 1914labour relations reform. It probably was institutionalized by thetripartite U.S. wartime regime.Introduction. . .51performance evaluations declined and although they continued to exist, they seldomhad a major impact on wage determination. Furthermore, in automobile plants, forexample, the wage gap between a production worker and a trades person becamefairly narrow with the hourly rate only about 20 percent higher for the latter.Job control rules refer to the web of contract clauses that determined thespecific tasks and rights of every employee. Job descriptions determined the contentof work and because different jobs had different pay rates, job switching was frownedupon. Seniority became a major factor in determining the outcome of bidding on jobs-- no longer did supervisors determine who would be posted where. Nor could theyarbitrarily decide which employees to layoff.7° As we shall discuss later, this notionof seniority, as a means of restricting employer discretion by limiting choice tomeasurable determinants such as length of service, is crucial in understanding why theterm “seniority-based wages” is so inappropriate from a comparative perspective indescribing Japan’s wage system. While managers in U.S. and Canada controlled thelabour process in theory, in practice the collective agreement regulated the regime andworkers were able to put an indelible, if incomplete, mark on the organization andregulation of work.Many of the points made above are generalizations and as such are subject towide variations depending on industry, region and workplace. But on the whole theyrepresent, in my opinion, a valid summary and an explicit starting point for° In the 1937 G.M.-U.A.W. agreement, clauses related to seniorityconstituted 35 percent of the contract, followed by grievance procedure at30 percent. The contract was still short at this time (186 lines oftypescript) . For further details see Gersuny & Kaufman, “Seniority andthe Moral Economy of U.S. Automobile Workers, 1934-1946.”Introduction. . .52comparing the practice of industrial relations and production management in Japanwith that in the United States and Canada. In particular, the de-centralized form ofindustrial unions, the rather egalitarian wage system, and the notion of seniority as acounter-weight to employer control are fundamental to any serious, comparativediscussion and will be referred to frequently as we proceed through the material.V. Case StudiesMany scholars would agree that Fordism, as defined by regulation theory is auniversal trend among industrialized countries (a form of convergence) although itsrealization differs according to country and the particular stage of economicdevelopment. In the case of Japan, it also adopted the social and economicdimensions of the Fordist paradigm but only after a period of extensive accumulationin the 1950s. The nature of Japan’s hegemonic regime can only be understood bytracing its origins back to the politics of production at the workplace.Three sites form the core of the primary research conducted for this study: theMiike coalmines (part of the Mitsui Coal group) in Kyushu, Suzuki Motors inHamamatsu, and Moriguchi City Hall just outside Osaka. These case studies werechosen partly through design and partly through good fortune. Miike was chosenspecifically because I had earlier studied the bitter, year-long dispute that occurredthere in 1960 and that became a landmark in Japan’s labour history. It seemed thatany solid interpretation of postwar period had to capture and explain the nation-wideconflict centred at Miike -- a conflict so broad in scope, so intense and so divergentIntroduction. . .53from the images oflabour-capitalharmony so oftenconjured up andoffered as regularfare for theWestern labourrelations specialist,that it could not bedismissed simplyas an exceptionthat proved therule. SuzukiMotors and Moriguchi City Hall became case studies more through fortuitouscircumstances than by design. I had, for other reasons, contact with these twoenterprises and thus they offered themselves as potential victims of scrutiny. Itshould be said, however, that together they conformed to one other explicit standardof my research goals, that is, diversity in the production process. Too often labourrelations models have been based exclusively on case studies in a single economicsector, thus depriving our reconstruct of the insights from other angles. In that sense,Moriguchi provided a window into the public sector workplace in Japan. SuzukiMotors offered the added attraction of being part of Japan’s automobile industry, aFigure 1.4: Location of Case StudiesMoriguchi City Hall(Moriguchi City)Coal(Omuta)Motors(Hamamatsu)Introduction. . .54sector which spawned much of the lean production paradigm being emulated today.The diversity of labour relations in these three sectors -- the mining, manufacturing,and public sectors -- affords a greater appreciation of the variation in industrialrelations that developed and continues to evolve in Japan today.On the whole, the research at these sites was focused mainly on workers andtheir unions. This work will hopefully become part of a growing body of scholarshipthat attempts to examine history “from the bottom up” as Harvey Kaye puts it.7’This tradition has long been upheld in Japan by scholars such as Yamamoto Kiyoshi,Totsuka Hideo, Hyodo Tsutomu, Hirai Yöichi, Matsuzaki Tadashi and KawanishiHirosuke who, in their voluminous works, have stressed the history of workers’struggles in Japan. It is only in the past decade, however, that works in English bysuch scholars as Andrew Gordon, Joe Moore, E. Patricia Tsurumi and NormaChalmers have begun to present this more balanced view of labour history or labour-management relations in Japan.72 They have, in their particular ways, “broughtworkers back in,” as Gordon puts it.While the case studies mentioned above form the core of this thesis, it is not. Harvey J. Kaye, The British Marxist Historians, (Cambridge, PolityPress, 1984)72 See Joe Moore, Japanese Workers and the Struggle for Power 1945-1947, (Madison, University of Wisconsin, 1983); Andrew Gordon, TheEvolution of Labour Relations in Postwar Japan (1853-1955) ; E. PatriciaTsurumi, Factory Girls: Women in the Thread Mills of Meii Japan,(Princeton University Press, 1990) ; Norma J. Chalmers, IndustrialRelations in Japan: The Peripheral Workforce, (London, Routledge, 1989)Moore has documented a period when workers and unions were setting theagenda in production politics in Japan. Tsurumi has given us a valuablecultural-feminist portrait of Meiji textile workers and Gordon hasattempted to show the workers’ role in the evolution of labour-managementrelations. Chalmers has provided an important contribution with heranalysis of the peripheral workforce in postwar Japan.Introduction.. .55restricted to them alone but ranges through a series of people, institutions andstruggles which I considered important to an understanding of postwar labour-management relations and hence to a different conceptualization of contemporary leanproduction systems. This has posed certain difficulties in the flow and structure ofthe narrative, but I believe the result is worth the extra trouble. Thus, apart from thethree detailed case studies, the thesis at times discusses more broad-scale events thatoccurred in the specific periods under discussion. In particular, I have tried tointegrate into the discussion the role of peak managerial organizations such asNikkeiren and labour organizations such as Sohyo. Given the significance of theautomobile industry, an effort was also made to compare Suzuki with Toyota Motorsin order to provide a broader perspective on the origins of lean production.Having said this, however, the reader should be cautioned -- this studyremains a very partial reconstruct of the postwar period and has its own limitations. Ihad hoped to include a case study based in textiles in which the labour force waspredominantly female. Predictably, this resolve came too late for extensive fieldwork and as a result, the current work affords only a cursory examination of thepolitics of gender. Nor is the peripheral work force, the 80 percent or more ofemployees who work in small and medium businesses examined in detail. Recently,other scholars have accorded this sector their attention but, given its magnitude, itremains woefully under-represented in our schema of production relations in Japan.73The methodology is basic. I have relied extensively on primary institutional. See N. Chalmers, Industrial Relations in Japan: The PeripheralWorkforce.Introduction. . .56histories by both management and labour for the case studies. This was supplementedby field work carried out at each site during a number of visits to Japan in the 1985-1990 period, including visits to the workplaces where possible and through numerousinterviews with rank-and-file workers, union officials and management representatives(see the bibliography for details). These visits also provided me with supplementary,first-hand accounts of important events. In addition, discussions with scholars ofJapan’s labour history have provided me with some scope and comparative referencepoints which I found very useful for finding my bearings in a very complex field.Introduction. . .57Chapter 2Testing the Limits: Workers’ Challenge(1945-1948)I. Revolution from the TopWhen Canadian diplomat Herbert Norman and U.S. foreign service officerJohn Emmerson headed for a prison on the outskirts of Tokyo in mid-October 1945,they carried with them authorization for the release of political prisoners held by theJapanese government. Effecting the release of two communists jailed for 18 and 19years respectively, two Korean independence leaders, anti-Fascist intellectuals, andreligious leaders was, reported Norman, “the most exciting experience of my life. “1This release, authorized by General Douglas MacArthur, Supreme Commander of theAllied Powers (SCAP), symbolized the onset of political revolution in Japan, arevolution whose liberal-democratic outcome was pre-determined because of theAmerican Occupation, but whose dynamics were shaped by fierce class conflict thatraged in Japan between 1945195O.2 The outcome of this conflict left an indelibleimprint on Japan’s workplace regimes as they evolved after the war. This chapter we1• As recounted in Roger Bowen, Innocence is Not Enough, The Life andDeath of Herbert Norman, (Vancouver, Douglas & McIntyre, 1986), pp. 118-119.2 In theory SCAP derived its authority from the Far EasternCommission that had 11 members states that included the Soviet Union.However, the postwar division of the world, worked out among the Allies,allowed the U.S., through SCAP, to dictate policy.Testing the Limits... .58examines the rise of a production regime that was shaped mainly by an insurgentlabour movement in 1946-1947.The political revolution launched by SCAP in the fall of 1945 createdpropitious conditions for the revival of the labour movement. In a detailed study ofAmerican Occupation policymaking, Michael Schaller concluded:During the initial reformist stage of the Occupation, lastingthrough early 1947, Washington encouraged SCAP to pursue aprogram that reflected the most progressive tendencies of the NewDeal. Even as American domestic and foreign policy lurched tothe Right, MacArthur and his aides remained committed to areform agenda abhorrent to most of the general’s conservativeconstituency in the United States.3To be sure the reform agenda had already been diluted by the American government’sdecision to retain the monarchy and to work through the established Japanesegovernment. Nevertheless, SCAP sponsored political and economic structuralchanges which would reverberate throughout society.4 On the political level thechanges included the proclamation of basic rights in October 1945 that released over3000 political prisoners; the passage of a Trade Union Law in December the sameyear; the drafting and passage of a new Constitution in November 1946 that investedsovereignty in the people, embraced a no-war clause, extended the franchise towomen, and articulated a Fordist social charter that guaranteed state assistance in. Michael Schaller, The American Occupation of Japan, the Origins ofthe Cold War in Asia, (New York, Oxford University Press, 1985), p. 25.. This overview on Occupation reforms is based on Michael Schaller,The American Occupation of Japan; John Dower, Empire and Aftermath(Cambridge, Harvard University Press, 1979); Howard B. Schonberger,Aftermath of War (Kent, Kent State University Press, 1989)Testing the Limits... .59welfare as well as the right to an education, labour rights and gender equality.SCAP also pursued a punitive course in arresting and trying suspected warcriminals (both military and civilian) and by “purging” from public office over200,000 people who allegedly helped direct Japan’s militarization. On the economiclevel, the Occupation began to dismantle Japanese industry as reparation payment inkind to Asian countries victimized by Japan. It implemented a trust-busting policyincluding obligatory stock sales to the public, anti-monopoly laws and an initialattempt to break up large companies. Land reform allowed tenants to purchase theirproperties and ended absentee landlordism.The purpose of citing these reforms is neither to tout nor to justify them butrather to sketch the early Occupation landscape. The fact is, the early anti-militaristand democratic tone of the Occupation put the prewar and wartime Japanese elite onthe defensive, if not in jail, and left them chafing at the bit of reform. Labour, onthe other hand, was liberated--free to organize and to play an independent role withinthe emerging liberal democratic structures of the day. The state, under the control ofthe U.S., had set new ground rules for society and for production politics.The early postwar situation was further complicated, however, by economicdislocation. The political structures may have conformed to an advanced capitalistcountry, but economically Japan was in turmoil. Although Allied bombing haddestroyed only 30 percent of Japan’s industrial capacity, actual production for 1946was down 70 percent from 1934-1936 levels. Tokyo and Osaka were bombed outwith nearly 60 percent of all buildings destroyed. Air attacks had decimated shippingTesting the Limits.... 60capacity. Rice production was seriously deteriorating. For city-dwellers starvationwas not only possible but imminent. Food shortages, a black market and excessivecurrency provoked a serious bout of inflation with prices doubling between October1945 and October 1946. The dislocation was further complicated by what may betermed a capital strike as major business leaders turned their backs on productionefforts. Why invest or produce when one might be jailed, purged or otherwisecompromised? Economic chaos, political freedom and a defensive capitalist classconspired to radicalize a working class that was in constant flux.Labour LawDemocratic reforms were imminent in the fall of 1945. On labour issues,however, Japan’s postwar elite was already one step ahead of MacArthur and, hopingto preempt potential liberal excesses, the Shidehara cabinet had authorized in Octoberthe formation of a tripartite commission to draft a trade union law. The labourrepresentatives on this commission were Matsuoka Komakichi and Nishio SuehirO ofthe pre-war JFL (Japan Federation of Labour, Nihon ROdO SOdOmei), MizutaniChOsaburo of the JSP (Japan Socialist Party, Nihon Shakai TO), Koizumi Hidekichi ofthe Japan Seamans’ Union (JSU, Kaiin Kumiai). Other members included MitsuiMining Company director Fukagawa Masao and University of Tokyo legal expertIzutaro Suehiro.5. RddO Sh ed., Shiry Rödö UndO Shi, 1945-1946 [Materials from theHistory of the Labour Movement] , (Tokyo, 1947), p. 689. This yearlyseries is hereafter abbreviated as SRUS.Testing the Limits.. . .61The committee quickly drafted labour legislation and on December 22, 1945the Diet passed the first Trade Union Law in Japan’s history. It contained fivesections and one addendum.6 Section I (Overview) declared that the purpose of thelaw was to stabilize the economy by elevating the status of workers and guaranteeingtheir rights to unionize and bargain collectively. Article 2 under this same sectiondefined employees and excluded from union membership employers and theirrepresentatives, organizations whose administrative expenses were provided byemployees, mutual aid societies and charitable groups, and groups whose mainpurposes were political or social.Section II (Unions) provided details of accreditation; unions had to havebylaws which were to be submitted to the appropriate administrative agency (gyOseikancho). Any disputes about union certification would be settled by theadministrative agency based on a recommendation by the Labour Relations Board.Unions were not liable for damages arising from reasonable actions during disputes.The courts could order a union to be dissolved for illegal activities upon applicationby the Labour Relations Board.Section III dealt with collective agreements which were to have a three yearlimit (Article 20); were to be based on a spirit of improving efficiency andmaintaining industrial peace (Article 21); were to be applied to all employees in caseswhere more than three quarters were under a collective agreement (Article 23); couldbe applied, at the discretion of the Labour Relations Board, to all workers in a given6 This analysis is based on the full text of the law as reprinted inROdO Shd, SRUS 1945-1945, pp. 771-774.Testing the Limits... .62industry of a specific region if a majority were already unionized (Article 24). Theparties could not resort to job action until after any mediation or arbitrationprocedures inscribed in the collective agreement had been exhausted (Article 25).Section IV outlined the provisions for Labour Relations Boards which were to betripartite (employer, union and public representatives on the board) in composition,and were to be established at both the central and regional levels (Article 26).The passage of the Trade Union Law affected labour relations both directlyand indirectly. On the one hand, it made unions legitimate both legally and socially.Furthermore, the creation of the Labour Relations Boards as the administrative agencyfor the law would itself create specific dynamics for the labour movement which willbecome evident as we examine the unionization process as it evolved at Miike, Suzukiand Moriguchi.A Re-Constituted Working ClassWhen the employees of Suzuki Looms, a textile manufacturer that converted toarms production during the war, gathered at the Hamamatsu head-quarters on the lastday of August 1945, there was little rejoicing. Suzuki Michio announced they wouldall be permanently laid off and thanked them for their years of service. Suzuki’smain facility and headquarters in Hamamatsu, Shizuoka prefecture, had been 95percent destroyed by American bombing. The glitter of gold was markedly absentfrom this final handshake for the nearly 2900 regular employees of whom nearly 1300were women. As for the 800 conscripted workers and 110 Korean prisoners-of-warTesting the Limits... .63that Suzuki had also employed during the war, there is no record.7 Yet within a fewdays of laying off all its employees, Suzuki rehired a select few, restarted operationsat its Takatsuka facility which remained unscathed, and in the last half of the financialyear (Oct. 1945-Mar. 1946), had sales of over 5.7 million yen.8As at Suzuki, there was a shakeout of the work force at the Miike mines inKyushu. Miike was the largest coal mine in Japan and had been owned and operatedby the Mitsui conglomerate since 1888. Over 24,000 miners were working Miike’sshafts at war’s end. But conditions at Miike were so bad, the miners abandoned thecoal seams en masse. By November 1945, 13,000 miners had abandoned the pits,including 6,000 Chinese, Korean and Caucasian prisoners of war.9 For these miners,defeat for Japan supposedly implied liberty, but while Caucasian prisoners-of-warwere quickly repatriated, Chinese and Korean miners at Miike had to riot before theywere able to leave the mines)0 In many cases, the Occupation policy was to forcenon-Caucasians to dig coal immediately after the war.11Coal production had been targeted as a priority industry early in theOccupation and as massive numbers of wartime workers abandoned the mines, the. Suzuki also employed over 100 Korean prisoners of war but nothinghas been recorded about their fate. See Suzuki Jiddsha KOgy, 40-Nen Shi[A History of 40 Years] , (Hamamatsu, 1960), p. 86.8 Suzuki Jidsha Kogyo, 40-Nen Shi, (Hamamatsu, 1960), p. 88.. Mitsui Kdzan Kabushiki Kaisha ed., ShiryO: Miike Sgi [The MiikeDispute: Materials], (Tokyo, Nihon Keleisha Dantai Renmei, 1963), p. 16.Ibid., p. 16.See Joe Moore, Japanese Workers and the Struggle for Power 1945-1947, (Madison, University of Wisconsin Press, 1983), pp. 33-35.Testing the Limits... .64coal operators were hard pressed to stabilize the labour force. Over 12,000 newminers were hired in 1946 but of that number nearly 8,000 quit or were dischargedwithin the year.’2 By December 1948, the workforce multiplied to reach its postwarpeak of 28,960. Of this total, however, only 1,000 had worked at Miike prior to1932.By comparison, Moriguchi employees did not experience the severe shakeoutin employment that workers at Suzuki and Miike had. Still not incorporated as a city,the few dozen employees in Moriguchi district remained at their post.II. Revolution at the WorkplaceThe Trade Union Law came into effect on March 31, 1946 by which time thelabour movement was already galvanized for action. The pace of organization wasfrantic and by 1948 over six million workers were in unions. To be sure, someunions were little more than paper organizations, but just as certain there were manyunions which were dynamic and member driven. As the labour movement confrontedemployers, it displayed a heterogenous mix of political proclivities, demands, andtactics, many of which failed to correspond to the norms that the Trade Union Lawhad hoped to establish. Further confounding the situation was the resistance tounionization displayed by employers. In some cases, workers and unions moved totake control of the workplace as it became increasingly evident that employers found12 Mitsui, Shiry: Miike Sögi, pp. 16-18.Testing the Limits... .65it difficult to break with their past paternalism, and proved intransigent in theirreluctance to move into any form of partnership with employees. In this confused andcomplex clash of interests, workers began to create new forms of workplaceregulation.MiikeAt Miike, management attempted to pre-empt the formation of an independentunion by creating a company union. Shortly after war’s end, Mitsui ordered newelections for Sanpo (the authoritarian wartime labour front) committeemen at eachmine site who then elected delegates to a union preparatory committee for eachFigure 2.1: Mitsui’s Milke Facifities(Omuta, Kyushu)Testing the Limits... .66site.’3 This process was well underway when miners at the Mikawa shaft (oneofthree main shafts at Miike, see Figure 2.1) rudely disrupted the processbydemanding an immediate 30 percent wage increase, a minimum wage and the eight-hour day.’4 As confrontation brewed, worker representatives undertook to form asingle union for Miike miners.On February 3, 1946, approximately 10,000 miners congregated in an Omutapark to found the Miike union. Also attending the gathering were city notables,management representatives and delegations from other unions. The union announcedits main demands: immediate participation in management; abolition of taxes onsalaries; large increases in severance pay; adequate money and materials to live on;the dismissal of foremen who stymied the workers will to produce; an end todiscrimination between staff and miners; severance pay for workers who quit; and anend to the black market.’5 The company gave each participant 10 yen for pocketmoney to show its good will toward the union. At this point the staff (shokuin) werenot included in the union.Mitsui’s reply to the union demands was piece-meal; it proposed the formationof a labour-management council in response to the union’s demand for participation inmanagement, but on monetary issues it equivocated. This led to a strike beginning on13 SanpO was the government-initiated and controlled Patriotic LabourFront that was modeled after the German wartime regime. It had committeesin every workplace.14 See Mitsui, Shiry: Miike Sgi, pp. 28-29 for details ofmanagement’s early attempts to control the union. According to Mitsui’saccount, prewar JFL organizers worked at Mikawa and they precipitated theearly demands and union formation.‘. Mitsui, Shiryd: Miike Sögi, p. 29.Testing the Limits... .67March 9 (the miners at Mikawa struck on Mar. 7). This action against Mitsui invitedlabour solidarity and on Mar. 4, unions at Tagawa and Yamano joined the Miikeunion in a joint council (Sanzan Kyogikai). The miners at these two Mitsui minespartially joined in the strike action. On the other hand, the Miike union decidedagainst occupying the mine and taking over production themselves (seisan kanri orproduction control) and condemned the Japan Communist Party (JCP) members whoappeared at the Mikawa shafts advocating such a takeover for their ‘inflammatoryattitude’ 16 The strike ended on March 13 with the company making some limitedconcessions. The co-ordination among the unions at the three Mitsui mines during theMarch strike led to the creation on May 19 of the WFMCU (Western Federation ofMitsui Coal Unions or Nishi Nihon Mitsui TankO ROdO Kumiai Rengökai), a regionalfederation of coalminers working at Mitsui’s three Kyushu mines.The following month, the Miike local signed its first collective agreement withMitsui. The contract was modelled after Section III (Collective Agreements) of theTrade Union Law but in addition gave the union important powers. The union wonformal recognition, the closed shop, a clause giving it a virtual veto over layoffs(article 3), the right to equal participation in a management council, and automaticrenewal of the collective agreement.’7By this time, however, workers in Japan were facing severe poverty and food16 Miike Tank Rd Kumiai Jü Nen Shi Hensan Iinkai, ed., Miike JüNen [Miike’s Ten Years], (Omuta, Miike Tankd Rddö Kumiai, 1956), p. 66.The entire contract has only eight clauses and was less than onepage in length. It is reprinted in its entirety in Miike Kumiai, Miike JIp. 58, appendix.Testing the Limits... .68shortages and the new collective agreement provided little in wage increases. AtMiike, miners were going into the mines without having eaten a meal and the regionalminers federation, the WFMCU, called for an immediate wage hike. Mitsui gave aone time bonus that averaged 250 yen per employee.’8The regional links developing among Mitsui mineworkers in Kyushu did notstop at the company gates. Although Tagawa and Yamano miners had left a regionalfederation that had included coalminers from other companies to take up affiliationwith Miike, this was not because of some predisposition towards ‘enterprise unionism’but rather because of political differences with that regional federation and because ofthe proximity with other Mitsui miners working in Kyushu.’9 In fact, the Mitsuiminers in Kyushu became quite active in helping found a ‘neutralist’ federation ofKyushu coalminers with workers drawn from a number of regional employers.The impulse for regional and industrial affiliations came from diverse sources.For example, on August 12, the acting director of the Central Labour Relations Boardmet with leaders of unions in the Omuta region and encouraged them to formindustry-wide affiliations.20 The WFMCU took the lead in establishing the KyushuFederation of Neutral Coalminers’ Unions (Kyushu Tankö Rödö Kumiai ChUritsuRenmei) on October 4. As evident in the name, this federation attempted to avoidpolitical affiliation with either the communist or social-democratic trends in the union. M±ike Kumiai, Milke Jü-Nen, p. 73.Shiryö: Milke Sogi, p. 31.20 Miike Kumiai, Miike Jü-Nen, p. 74.Testing the Limits... .69movement.21Parallel to this political polarization, an attempt was made to overcome theresultant schisms through the formation of an umbrella group which would allow aunited front against the employers. On the regional level this manifested itself in theformation of the Fukuoka Prefectural Council of Coalminers’ Unions (Fukuoka-KenTankO Rödö Kumiai Kyogikai) on October 21 (see Figure 2.2). On the national level,a similar umbrella group loosely affiliating the three trends, the Japan Coalminers’Council (TankOROdO Kumiai Figure 2.2: Milke Union’s AffiliationsZenkoku KyOgikaiLocal/Enterprise Levelor Tankyo) was Miike local Western Fed. Mitsui(Feb., 1946> -> of Mitsui Coal -> Mineworkersestablished on Unions Federation(May, 1946) (March, 1949)January 25, 1947. PrefecturalFukuoka Prefectural Council of340,000 miners or Coalminers Unions(October, 1946)83.4 percent of the KyushuKyushu Federationcoal labour force of Neutral Coal-Miners’ Unions(October, 1946)was represented inNationalthis federation Japan Coalminers’ Japan Coalminers’Council Union (JCU)(January, 1 947) (October, 1947)which begannegotiating with21 This organization later changed the “Neutral” component of itsname to “Democratic” after being criticized for attempting to stake out amiddle ground between employers and employees. See Miike Kumiai, MiikeJD-Nen, pp. 74-75.Testing the Limits... .70the national coal employers group immediately. At this time, the union advanced aproposal for a social wage; it demanded a sliding wage that would allow the miners topurchase a daily quota of 2400 calories. The interim agreement signed on March 7only went half way to meet the demand and all hell broke loose as the left-leaningunions wildcatted in protest. This led to the demise of the regional and nationalumbrella federations. The right and centre federations consequently consolidated andunited to form the JCU (Japan Coalminers’ Union; Nihon Tankö ROdO Kumiai DOmeior Tanro) in October 1947 and the left-leaning unions organized the All-Japan CoalFederation (Zen Nihon Sekitan Sangyo ROdö Kumiai or Zen Sekitan).22A number of significant points emerge from this brief sketch of unionorganization and conflict in the coal mines. Miners at Miike evidently wanted anorganization that was independent of company control; politics played an importantrole in the alignment, disintegration and re-alignment of union federations; regardlessof political stripe, miners wanted affiliations with other miners across the enterprise,both regionally and nationally. This natural inclination to develop horizontal linkageswith other miners was given further impetus by the call from the Labour RelationsBoard encouraging the formation of cross union linkages. The coal unions also tookup the demand for a social wage. In a sense this was hardly surprising. For workersin Japan to accept that wages should be pegged to an almost non-functioning marketwould have meant starvation in the immediate postwar period. Politics and economicsconverged, propelling many unions to conceive of workers’ interests independent22 Nihon TankO RödO Kumiai DOmei ed., Tanro Ju Nen SM [A History ofTen Years], (Tokyo, RädO JunpO Sha, 1961), p. 162.Testing the Limits... .71from the state of the market.SuzukiThe tendency towards independent unionism was as spontaneous as it wasorganized. Suzuki workers also formed a union in early 1946. After havingdismissed all its employees at war’s end, the company proceeded to hire back 350workers. On January 18, these employees turned around and formed the SuzukiLoom Workers Union. Documentation regarding the original union is scanty butaccording to one source, the inspiration behind the union came from amongsupervisors 23 This notwithstanding, horizontal affiliations began immediately as theunion was one of the founding affiliates of a regional federation, (EishU ChihO ROdOKumiai Kaigi, Eishu Regional Labour Federation, founded on February 25) whichundertook to co-ordinate union activities for May Day and to deal with the sharpeningfood crisis that had begun to grip the country. The president of the Suzuki union,Kawai Kasaku, became a leader of the regional federation.As in the coalminers case, the impetus for horizontal links amongmetalworkers was the Trade Union Law. Promulgated on March 1, 1946, the newlaw called for the creation of regional labour boards composed of representatives of23 According to Sugiura Kiyoshi, one of the founding members of theunion as cited in Suzuki Jiddsha KOgyd Rddö Kumiai Shi Henshü Iinkai ed.,Niü Go-Nen Shi [A History of Twenty-Five Years] , (Hamana-Gun, SuzukiJiddsha Kgyd Rödd Kumiai, 1976), p. 225.24 Shizuoka Ken ROdO UndO Shi Hensan Iinkai ed., Shizuoka Ken ROdOUndO Shi [A History of the Labour Movement in Shizuoka Prefecture],(Shizuoka, Shizuoka Ken ROdO Kumiai HyOgikai, 1984), p. 292-295.Testing the Limits... .72employers, labour and the public. The Shizuoka labour office (Shizuoka KenRoseika) subsequently pushed for unions to join together to choose prefecturalrepresentatives.25 Thus, on Feb. 6, union representatives from across the prefecture,including Suzuki’s Kawai, gathered at the first meeting of the Preparatory Committeeof the Shizuoka Labour Federation (Shizuoka Ken ROdO Kumiai RengOkai Junbi Kai).The meeting not only designated labour representatives to the labour board but alsobegan deliberations for a permanent labour federation which would co-ordinateactivities for both industrial and regional unions. On March 2, 1946, the ShizuokaLabour Council was founded with Kawai elected as vice-president. The Suzuki unionthus developed strong regional links early in 1946. Industry-based affiliations alsobegan in this period. In Shizuoka, unions from the metal industries began meeting inMay 1946, and in October the Japan Machine Tools Union-Shizuoka District wasfounded with the Suzuki-Style Loom Union affiliated as one local.26Management at Suzuki was adamantly opposed to the union’s affiliation to theleft-leaning Machine Tool Union and this led to the first postwar confrontation at theplant. In November 1947, the union local demanded that Suzuki conclude a newcollective agreement and recognize the new union. Suzuki declined but did begin to‘consult’ (kyogi suru) on the issues of wages and year-end bonus. Internalconciliation efforts failed and on January 8, 1948 the union struck for 24 hours. Onthe same day local union members crashed a meeting between union representatives25 Ibid., p. 277.26 Shizuoka Ken R5dö Und5 Shi, p. 296-298.Testing the Limits... .73and Suzuki Michio. Barring all exits, the militant unionists forced Suzuki to bargainfrom 4 p.m. on the 8th until 3:30 the next morning when a tentative agreement wasreached. This agreement fell through prior to the official signing on the afternoon ofthe 9th and the union immediately occupied the plant and entered into productioncontrol. Suzuki finally concluded a new wage agreement on January 15, raising thebase wage to 3800 yen per month with an average 2,000 yen year-end bonus.In August 1948, Suzuki agreed to sign a new collective agreement thatincluded provisions for a union shop (Article 4), a union veto over hiring and firing(Article 5) and changes in compensation and personnel (Article 15), as well as theright to automatic renewal of the collective agreement should a new agreement not beconcluded prior to the expiration of the existing agreement (Article 22) and equalparticipation in a managerial council (keiei kyogikai).27The nature and achievements of the Suzuki union, with its growing regionaland industrial affiliations, its gains in employment security and a strong say at theworkplace, were strikingly similar to those of the Miike miners. These similaritiescannot be attributed to a common leadership since the two unions were affiliated withdifferent labour federations, both regionally and industrially. Union politics alsodiffered, with the Miike union taking a distinctly neutralist tone while the Suzukiunion was moving into the left camp.In both cases, the unions took on a distinctively independent flavour and27 The entire collective agreement is reproduced in Shizuoka Ken ROdOUndd Shi Hensan Iinkai ed., Shizuoka Ken RdO UndO Shi, Shiryd (Ka)[Documents, Vol. I: The History of the Shizuoka Labour Movement](Shizuoka, Shizuoka Ken ROdO Kumiai HyOgikai, 1981), pp. 399-402.Testing the Limits.. . .74developed affiliations with workers on both a regional and industrial basis. There wasno stopping at the company gates. Early achievements seemed to reflect aspontaneous drive for secure, independent organizations to represent them in dealingswith management, for secure jobs and for a say at the workplace. In other words,unions reflected workers’ distrust of the situation and their unwillingness to letemployers sort it out. The Trade Union Law and its administrative organs acted asboth guide and at times as catalyst, channelling the spontaneous movement intodistinct directions.The Moriguchi City UnionMoriguchi was incorporated as a city in November 1946 through theamalgamation of two local districts. According to the city workers’ union history, theunion’s beginnings can be traced to a dispute related to wage discrimination betweenemployees of the two districts.28 Kiyomizu Yasuji, director of the economicsdepartment, led a group of 20 disenchanted employees and confronted city officialsover the discrimination in pay. Subsequently, on December 12, Kiyomizu and othersmet in a local school to found a union to represent all city hall employees (includingsupervisors and foremen). Little has been recorded about the union’s activities in thisearly period except that the Moriguchi union did develop horizontal affiliations with28 Morigichi-Shi Shokuin Rödö Kumiai ed., Moriguchi-Shi ShokuröSan-jU Go-Nen Shi [Thirty Five Years of the Moriguchi Employees Union](Moriguchi, 1981), p. 82-83.Testing the Limits... .75other public sector unions in the Osaka region, probably in 1948.29The propensity for horizontal linkages was one of the outstandingcharacteristics of unions at all three worksites. It was also a distinctive feature formost unions at this time. Part of the reason for this was a spontaneous desire forrelations with other unions. Regional and industrial affiliations were given furtherimpetus by the Trade Union Law. Politics also played a role as union activists ofvarious political persuasion attempted to organize on a national scale in an effort toeffect change at levels beyond the enterprise.National Labour Federations and the Winter of DiscontentIn this early period, the Miike, Moriguchi and Suzuki local unions had bothregional and industrial affiliations, while the Suzuki union was also affiliated with anational labour central. Through its affiliation with the national Machine Tool Union,it became part of the National Congress of Industrial Unions. The first impulse forthis labour central had come from the newspaper unions and the KantO labourfederation. On February 20, 1946, they founded the Preparatory Committee for aNational Congress of Industrial Unions (NCIU, Zen Nihon Sangyo Betsu ROdOKumi