Suppliers' associations in the Japaneseautomobile industry: collective action fortechnology diffusion
Mari Sako*
This paper analyses the structure and functions of suppliers' associations (kyory-okukai) in the automobile industry in Japan. The bilateral assembler-supplierrelationship has received much attention recently as a source of Japanese industrialcompetitiveness. However, this paper argues that the hitherto neglected area ofintersupplier co-ordination in technology diffusion is at least as important as thebilateral assembler-supplier relationship in accounting for the overall performanceof the Japanese automotive industry. On the basis of company visits and alarge-scale survey of first-tier suppliers conducted by the author, the paper analysesthe reasons why suppliers' associations were established and why they continue toexist today, and their effects on economic performance.
The Japanese economy is often described as a 'network economy', with some distinctionsbeing made between enterprise groups and networks (Imai, 1994), or between inter-market and vertical keiretsu (Gerlach, 1992). A decade ago, when Italian industrialdistricts were rediscovered as a source of innovation and local economic regeneration(Piore and Sabel, 1984), scholars looked for equivalents in Japan. The same principlesunderlying flexible specialisation were found to be operating in large, decentralisedmultinational firms (Sabel, 1989) as in an engineering district of Sakaki (Friedman,1988) (which, incidentally, is just one of the 549 industrial districts in Japan; SMEA,1989, p. 117).
The suppliers' association (known generically as kyoryokukai), die focus of diis paper,is a highly relevant type of network in this context. Such associations exist at all 11Japanese vehicle manufacturers except Honda. In addition, over 300 primary partssuppliers, many of which are members of vehicle manufacturers' associations, have theirown associations of suppliers (Dodwell, 1986, cited in Smitka, 1991). The suppliers'association is therefore an institution widi a significant presence in the Japaneseautomotive industry.
Manuscript received 6 March 1995; final version received 29 August 1995.
*London School of Economics and Political Science. I should like to thank the anonymous referees andSudipto Bhattacharya, Ron Dore, Susan Helper, Richard Lamming, Ray Richardson, Michael Smitka,Kazuo Wada, Hugh Whittaker, Stephen Wood and Hideki Yamawaki for providing helpful comments onearlier drafts. The International Motor Vehicle Program (IMVP)'s financial grant and JAPIA's support forconducting the supplier survey reported in this paper are gratefully acknowledged. I should also like to thankJAMA for arranging visits to all the 11 assembler companies in Japan.
Despite this, suppliers' associations, their origins and their raison d'etre have beenrelatively understudied in recent years because the main paradigm for analysing theleading sector of the Japanese economy does not allow for a network-like analysis.Traditionally, the suppliers' association featured in scholarly inquiries among JapaneseMarxist economists as a tangible institution for large firms (monopoly capital) to exerciseunilateral control over their smaller subcontracting firms. More recendy, however, theMarxist paradigm has been superseded by transaction cost economics and game theory,which conceives transactions as decomposable into bilateral contracts (Williamson,1984; Grossman and Hart, 1986). In this framework, the Japanese car industry has acompetitive edge because the assembler has forged a long-term and recurrent relation-ship with each of its core suppliers. The stability of die assembler-supplier relationshipenables the supplier to contribute to design and development, to make investments andto accumulate know-how that may be useful only to that relationship. Such 'relationship-specific skills' (Asanuma, 1989) are a major source of superior performance. Thisanalysis of the bilateral relationship has tended to underestimate the significance ofhorizontal co-ordination among suppliers. Intersupplier rivalry is certainly promoted bysome Japanese assemblers' practice of making public the ranking of dieir core suppliersaccording to performance (see Wada, 1991, p. 9 for Toyota). Relative ranking creates anincentive to engage in continuous improvement (Aoki, 1988). However, this techniqueof managing die supply chain can be implemented without a suppliers' association.
The above paradigm may, in part, account for a view diat the suppliers' association isa redundant institution whose existence has far surpassed its utility. Until the 1970s,most assemblers were engaged in disseminating technical and organisational practicessuch as total quality control (TQC), value analysis or engineering (VA. VE), and'Just-in-Time' (JIT). The suppliers' association was a convenient forum for providingtechnical assistance en masse, while minimising the spill-over of benefits from suchassistance to competing assemblers. There were, dierefore, significant 'association-specific rents' as well as bilateral relationship-specific rents. However, with slower andless assured growdi in the 1970s and 1980s, more suppliers started to diversity their riskby trading with several assemblers. Assemblers began to undertake less top-downtechnical assistance as benefits from association activities could no longer be madeexclusive to one assembler, and more technologically capable suppliers emerged overtime. According to diis view, suppliers' associations continue to exist out of inertia. Theyare more like social clubs and do not contribute much to the overall efficiency of theJapanese automobile industry today.
A view that goes beyond asserting the declining utility of suppliers' associations isbased on die Adam Smidiian notion of a businessmen's conspiracy against die publicinterest. This was most recendy expressed during die US—Japan Structural ImpedimentsInitiative (SIT) discussions. Here, the suppliers' association was cited, along widihorizontal enterprise groups, as keiretsu, a group of firms like a cartel diat exists to protectits monopoly profit by excluding outsiders. According to diis view, die suppliers'association constitutes an unfair trading barrier that ought to be dismanded. Suchcriticism led Nissan, in 1991, to merge its two associations (Takarakai and Hoshokai)into one (Nisshokai), to which non-Japanese suppliers are increasingly admitted. Also,most of die Japanese assemblers have endeavoured to make the criteria for becomingassociation members more explicit and transparent dian in the past. However, theJapanese government's response, and its official line spelt out in recent Economic WhitePapers (EPA, 1990, p. 196ff; EPA, 1992, p. 276ff)3 have been in terms of the bilateral
Suppliers' associations 653
relationship paradigm only. They have thus far failed to investigate whether the suppliers'association itself has any significant economic impact or not.
From the above, it is clear that a study of suppliers' associations is needed in order toclarify how they function and what their effects are. This is in anticipation of an increasedinterest in suppliers' associations from the following angles. First, keiretsu critics wouldwish to know whether the associations are indeed exclusionary with strict boundaries,and whether they ought to be dismantled. Second, as more non-Japanese firms areadmitted into suppliers' associations in Japan, potential overseas suppliers would need tohave a good understanding of what association membership entails.1 Third, is thesuppliers' association a method of managing suppliers that would work only in Japan (asWomack et al., 1990 seem to imply)? Alternatively, is it an organisational form worthy ofemulation by Japanese and non-Japanese assemblers located in North America andEurope?2 Or is it a historical anachronism that would eventually wither away with theglobalisation of the car industry?
The evidence presented in this paper does not support the prevailing views in theliterature spelt out above. It is shown (a) that contrary to the bilateral contracting view,suppliers value mutual learning from other suppliers just as much as learning from theirassembler-customer, (b) that the majority of first-tier suppliers do not consider suppliers'associations to be of less use now that in the past, and (c) that association members havelower pre-tax profitability than non-members, a piece of evidence that undermines theview of the association as a cartel-like entity. The empirical analysis will be based on datacollected by the author through (a) company visits and interviews of purchasingdepartments and suppliers' association offices of all Japanese assemblers and somefirst-tier suppliers in 1992 and 1993, and (b) a large-scale survey of first-tier suppliersconducted by the author in July 1993.
This paper is structured as follows. The first section provides an overview of themembership structure, growth and turnover. This is followed by a brief account of thehistorical development in Section 2. Section 3 examines the contemporary functions ofthe suppliers' association and their effects.
1. The suppliers' association as a Japanese business networkConsiderable variations exist from association to association in their characteristics.First, as shown in Table 1, the number of member suppliers per assembler varied fromas few as 97 at Suzuki to 362 at Mitsubishi Motors. However, relative to assemblers'production levels, Toyota has a concentrated membership (17,839 vehicles produced perannum per member supplier) while smaller assemblers have dispersed membership (only379 vehicles per member supplier in the case of Hino). A relatively large assembler,Mitsubishi Motors, also has a dispersed membership, with 3,883 vehicles produced permember supplier.
1 As of September 1992, 18 members of Mitsubishi's Kashiwakai were non-Japanese, including Bosch,GM Japan Allison, and Simpson Industries (Nikkan Kogyo Shinbun, 11 September 1992). Nissan'sNisshokai had 9 non-Japanese members, including Goodyear Tire and Rubber (Nikkei Sangyo Shinbun,4 August 1992).
2 Hines (1994) reports the case of LJaneUi Radiators Supplier Association, which was established with theadvice of the Welsh Development Agency. The UK Department of Trade and Industry is also providingfinancial help to set up a suppliers' association for Rover. Such public sector consultancy and support wereindispensable in Japan also, at the time when supplier associations were still in their nascent stage in the1950s (see Section 3 of this paper). Also, Andersen Consulting (1993) recommends that the topmanagement of European companies 'build shared destiny relationships with your suppliers—throughestablishing supplier clubs' (p. 5). See also Lamming (1993).
654 M. Sato
Table 1. Membership of suppliers' associations, 1990
Auto makers'association
ToyotaTokai KyohokaiKanto KyohokaiKansai Kyohokai
NissanTakarakaiShohokai
MitsubishiKashiwakai
Mazda YokokaiWest JapanKantoKansai
IsuzuKyowakai
Fuji Heavy IndustriesSubaru Yuhikai
DaihatsuKyoyukai
Suzuki MotorsCooperative
HinoSuppliers' association
Nissan DieselYayoikai
Honda**
Foundingyear
1939/194319471947
19541966
19711981
1962
1986
1969
1957
1962
1960
n.a.
Totalmembers
(A)
229*1416325
174*104
70262
180*626256
284
218
168
97
238
57
297
Vehicleproduction(units) (B)
4,085,081
2,330,943
1,405,647
1,385,941
470,950
528,333
670,481
858,268
90,269
60,529
1,358,415
Ratio(B)/(A)
17,839
13,396
3883
7700
1658
2424
3991
3693
379
1062
4574
*Some mulriplant supplier companies may belong to more than one of theregionally based associations of an assembler. The numbers shown in this tableinclude such double-counting.
**Honda does not have any association. The suppliers refer to those listed as'main suppliers' in the first source.
Sources: Auto Trade Journal and JAPIA (1990 Nihon nojidosha Buhin Kogyo;Japan Automobile Manufacturers Association (1991) Nihon nojidosha Kogyo.
Perhaps the most significant reason for the varying degree of concentration ordispersion of membership lies in the location of assemblers' plants. At one extreme,Toyota had, until very recently, located all of its plants in and around Toyota City, wherea majority of its suppliers are also located. Toyota plants in the same locality have sharedtheir supplier base. At the other extreme, Mitsubishi Motors have plants that aregeographically dispersed, in Tokyo, Nagoya, Kyoto and Mizushima. Each developed itsown local supplier network. Even after the four plants were incorporated into the newlyindependent Mitsubishi Motor Corporation in 1970, the scope for consolidating thesupplier base was limited. This limitation was aggravated by Mitsubishi's wide productvariety, ranging from large trucks and buses to small passenger cars.
Automotive suppliers may be broadly classified into three categories: part suppliers,raw materials suppliers and suppliers of equipment and tools. Most assemblers havean association exclusively for parts suppliers, while some assemblers have separate
Suppliers' associations 655
Table 2. Members of suppliers' associations as a proportion of total suppliers, 1992
P, Parts suppliers; M, materials suppliers; E, equipment (including tool and jig) suppliers.Source: Company visits and interviews by the author in 1992 and 1993.
associations for different types of supplier. This paper concentrates on die associations ofparts suppliers as they are numerically the most significant. In general, associationmembers take up a large proportion—around 80% or 90%—of each assembler'sexpenditure on purchasing parts (see Table 2). However, the degree to which theassociation is encompassing varies from assembler to assembler. The most encompassingwas Kyohokai, whose 183 members, out of a total of 350 parts suppliers, accounted fornearly all (98%) of Toyota's total purchasing expenditure on parts. At the other extremewas Suzuki's association whose 105 members accounted for only 31% of die totalspending on parts.
Next, suppliers' associations are by no means congruent with the keiretsu group, whichmay be denned as a self-defined business community bound by virtue of havingoverlapping and multiplex ties in shareholding, personnel and trading (Gerlach, 1992).For instance, in the case of Toyota's Kyohokai association, members may be classifiedbroadly into three categories. First, there are 10 member companies that Toyota itselfdefines as part of die so-called Toyota Group (for example, Nippondenso, Toyota AutoBody and Aisin Seiki). The Group firms are linked through a complex reciprocalshareholding pattern, with around a quarter of the Group firms' total shares being heldwithin the Group. Besides the 10 Group companies, there are 25 Kyohokai memberswhose largest shareholder is Toyota Motors. Thus, around a fifth of the total 183Kyohokai members are part of the core vertical keiretsu group.
Second, there are around 40 locally based sub-contractors that tend to be independentin shareholding and personnel aspects, but are heavily dependent on orders placed byToyota (Ueda, 1989, pp. 15-16). Their businesses are mainly in mechanical engineeringsuch as metal pressing, casting and forging. They have a long history of trading withToyota, from which they have received technical and managerial assistance. If unions arerecognised at these firms, they tend to belong to the All Toyota Federation of EnterpriseUnions. These links, both managerial and labour, may warrant grouping these firms aspart of the Toyota keiretsu group.
This, however, leaves us with just over a half of Kyohokai members in the thirdcategory, consisting of mainly medium-sized independent firms that do not owe
656 M. Sako
allegiance to any particular assembler for their origin and growth (such as AkebonoBrake), and some large corporations (such as Toshiba and Dunlop Japan). Theseso-called independents constitute a non-neglibible proportion of the other assemblers'association membership. Some assemblers' association members are easily categorisableinto relatively dependent sub-contractors and larger independent suppliers. For example,Nissan, before the 1991 reorganisation of its association, had made the clearestdistinction by having two separate organisations. Takarakai with 104 members in 1990was for smaller suppliers who tended to be heavily dependent on Nissan's business, andShohokai with 70 members was for larger independent suppliers.
Once it is established that suppliers' association membership emcompasses groupingswhich are broader than keiretsu groupings, it is perhaps not surprising to discoversignificant overlaps in membership. Even arch rivals, Toyota and Nissan, shared 44suppliers in their respective associations in 1985 (Ueda, 1989, p. 11). Moreover, this isnot just a recent phenomenon, as 32 suppliers belonged to both Toyota's association andone of Nissan's associations as early as in 1967 (Miwa, 1990). In 1992 there were 191members of Nissan's Nisshokai, of which 55 were also members of Toyota's Kyohokai(the author's calculation based on Auto Trade Journal and JAPIA 1993). Thesecompanies were predominantly manufacturers of tyres and rubber parts, glass, paint,batteries, electronic parts, bearings and brake systems. By contrast, suppliers which haveremained members of a single association tended to be locally based single-establishmentcompanies in mechanical engineering.
Over time, there has been a growth in multiple membership of suppliers' associations.In particular, the number of parts suppliers that participated in five or more suppliers'associations increased from 67 in 1980, to 81 in 1985 and 93 in 1990 (see Table 3). Ofthe 93 in 1990, 21 firms were members of eight major associations (i.e. Toyota, Nissan,Mitsubishi, Mazda, Isuzu, Fuji, Daihatsu and Hino) and were also main suppliers toHonda. These companies employ around 3500 workers on average, and produce brakesystems, bearings, springs, spark plugs, tyres, belting and batteries. They tend to play aleading role in association activities, by taking on chairmanship and executive positionsin association committees and meetings. Thus, instead of regarding the automotiveindustry as consisting of 11 overlapping keiretsu groups each headed by an assembler, theindustry may be better characterised as a network, with the 20 or so core primarysuppliers transmitting information from one association to another. Assemblers witha relatively new suppliers' association, such as Mazda, have learnt the know-howin running the association from these core suppliers rather dian from competitorassemblers.
Why has multiple membership of suppliers' associations increased over time? Possiblecandidates for major causes of multiple membership are: (a) slower growth after the 1973oil crisis, which led assemblers to insist less on exclusive supply; (b) the electronificationof automotive technology, which led assemblers to source from non-traditional sources(Ikeda, 1989) and from sources possessing technological expertise that the assembler hadlittle hope of being able to match in the short run; and (c) the globalisation of theJapanese automotive industry, and especially foreign direct investment by Japaneseassemblers that sought new sources of parts supply. For example, in 1989, Nissan startedpurchasing electric fuel pumps from Nippondenso for use at Nissan Motor Manufac-turing Co. in the USA (Nikkei Shinbun, 4 September 1990). In 1990 Tachi-S, aNissan-affiliated seat manufacturer, obtained an order for Toyota's new compact carmodel (Nikkei Shinbun, 15 January 1990). Similarly, Hitachi, a Nissan supplier, started
Suppliers' associations 657
trading with Toyota {Nikkei Shmbun, 9 February 1993), while Zexel, whose majorbusiness has been with Nissan and Isuzu, also stated supplying Toyota. However, neitherHitachi nor Zexel nor Tachi-S is a member of Toyota's Kyohokai. Nor is Nippondensoa member of Nissan's Nisshokai. Trading embodying strategic technology is evidentlypossible wimout association membership.
The trend towards overlapping membership was accompanied by an increase in thetotal number of members over time. Such increases were conspicuous at smallermanufacturers such as Mazda, Fuji and Isuzu, which tried to tap into the morecompetitive supplier network of Toyota and Nissan.
Small changes in the total number of association members, of course, do not precludehigh turnover, with new entrants replacing those diat exit (Table 4). In the 1970sNissan's associations had the lowest turnover but, by the 1980s, Toyota's associationemerged as the one with the lowest rates of quits and entry. At Toyota, the TokaiKyohokai had 105 members in 1963. By 1971 there were 120 members; only 5 of the oldmembers had been dropped, while 20 were added (Smitka, 1991, p. 85). During1971-1981, 21 new suppliers joined, while only 3 left; and during 1981-1992, 8 joinedwhile 2 left the Tokai Kyohokai.
By contrast, at Mitsubishi Motors, 84 firms entered while 83 firms left the Kashiwakaiduring 1971-1981, and 73 entered while 46 left during 1981-1990. Thus, on a roughcount, of the 358 members in 1990, only about a half remained members over the entire20-year period (see also Smitka, 1991, pp. 85-87).
What factors account for the differences in turnover rates among associations? Onereason appears to be differences in the assemblers' product strategy. In the Japaneseautomobile industry, it is well known that the implicit supplier contract is for theduration of a model cycle. This implies that the possible occasions for the assembler toswitch suppliers, and hence for potential entry and exit of association members, are morenumerous the greater the product variety and the shorter the model cycle. Given that thelength of the model cycle is more or less the same across assemblers, the greater varietyof vehicles manufactured by Mitsubishi Motors, compared with Toyota, perhapsaccounts in part for the higher turnover rate in Mitsubishi's Kashiwakai membershipthan in Toyota's.
Product or marketing strategy may affect the scope for continuous sourcing in anotherway. In particular, a contrast may be drawn between two broad types of marketing. Onthe one hand, some assemblers, such as Toyota, pursue full-line marketing with anemphasis on the continuum in the spectrum of models from low to high price. On theother hand, other assemblers, such as Honda and Nissan to an extent, pursue asegmented market strategy with an emphasis on bringing out discrete 'hits' targeted atspecific customer groups (Itami et al., 1988, ch. 5). The former can take better advantageof common styling and parts over model cycles as well as across existing models than canthe latter. Therefore, supplier relationships can be expected to be more continuous atfull-line strategy assemblers like Toyota than at segmented strategy assemblers likeHonda.
To summarise, there is a considerable variation from association to association withrespect to (a) the size of membership, (b) the proportion of members in the total supplierbase, and (c) turnover of members over time. However, as a common characteristic,association membership is much broader than the boundary of keiretsu groupings,particularly in recent decades when independent suppliers which stand outside thekeiretsu have been taking up membership in multiple associations.
658 M. Sako
Table 3. Multiple membership in supplier associations 1975-1990
2. Historical origins and contemporary context
The suppliers' association (kyoryokukai) literally translates as a 'co-operation associ-ation'. It is a voluntary association with its own rules and regulations. Its aim is generallysaid to be to enhance member suppliers' co-operation with the assembler and with eachother. Most of the suppliers' associations have a name that signifies co-operation,friendship or prosperity. Some associations, just like Japanese companies, are described
Suppliers' associations 659
Table 3. Continued front previous page
Toyo RubberToyo Radiator
Toyoda GoseiNiles Parts
NifcoNishikawa Rub.Nisshin KogyoNittan Valve
Nippon St. Glas1 Nippon Air Br
Nippon Cable S.Nippon Seiko
Duntop JapanNippondenso
Jap. Stor. Batt.Nihon Power S.NHK SpringNip. Piston R.Nihon MichelinBando Chem
• _1 H _•• BBriBH BHH m Ml 111 HJ1; 111 BHH rm • H • MM•fl • 11 m H WM H
iI
1BH ffl W ' • M • •BH Ml I 1 HI HI HM "1 H 11V M M1 1 ] 1 1 E I BM M ; M I ! i | | l | |• • •BH ^ BHMl I H Mfl
• Ml HI BH• MM fW MlmrTTi I M m
jH_^_BijXm_mL^B_M_ffn M I n™1
M rm• • •• • •MMHJ , M^J i L^J
j M J H H M J H H M J I ' l l
J H J _IIRH• • • i
• • • \\\ HI HI . 1
m M M Ml, m in I B M 1 BM BBI 1!!BH 1 ^MH ™ H• • •• • •in
1MB jWIMjn BH nn w;'^•| |8 y• • • IHJ M m HI ' 'B l B*J ' i 1 ^^1 :
B l ^Bl i i 1 B*J
Hino | ND | Honda
•H M
1 1 BBI
m 1
BHBH-Hmt 1
BL_MMTMM^BHJHX ;_BH •-HI - - --
M l
Hi
1
Nil
•BJi i |
11B H
ru•
MTOMHI
1 1 Bl^Hl ' 1 MlBBI 1 BBIHi H I
Sote: Shades indicate membership in that year. Key: • 1975, D 1980, Q '985, • 1990.
as a 'community of fate' (unmet kyodotai). Tracing the historical origins and the evolutionof suppliers' association assists us in understanding these sentiments.
2.1. Historical evolutionThe oldest of the supplier associations is Toyota's Kyohokai, which may be traced backto a gathering in 1939 (Kyohokai, 1967, p. 10), although a formal association was notlaunched until 1943. As part of the wartime control regime, the Japanese government atthe time imposed a regulation for nominating small and medium-sized firms to supply to
660 M. Sako
Table 4. Entrants and leaven of suppliers' associations, 1971-1990
Suppliers'
Toyota
Nissan
Mitsubishi
Mazda
Isuzu
Fuji
Daihatsu
Suzuki
Hino
NissanDieselHonda
association
TotalTurnover*
TotalTurnover
TotalTurnover
TotalTurnover
TotalTurnover
TotalTurnover
TotalTurnover
TotalTurnover
TotalTurnover
TotalTurnover
TotalTurnover
1971/2-1980/1
Entrants
472 1161 0842-5443-4963-9955 1332-2424-6
1326-512
2 0n.a.
Leavers
391-714
0-9832-5594-5331-3432-3503-4182 0432 117
2-8n.a.
1981/2-1989/90
Entrants
2010302-0733-3221-4491-9804 1322-319
2-1211-0
51-0
1956-7
Leavers
130-6151 0462 1171 1441-7763-9
70-5243 0271-3
71-4
2398-2
Thanks to Toshiko Iwase and Arisa Ouchi for assistance in compiling thistable.
*Average turnover rate per annum=Average number of entrants or leavers perannum/[(Membcrs in initials year+Members in end year)/2].
Source: Author's own calculations based on the data in Auto Trade Journaland JAPIA Nihon no Jidosha Buhin Kogyo, various years.
large firms in order to control industrial production for the war effort. Non-designatedfirms were left to perish owing to lack of funds and materials (Nakamura, 1986,pp. 124-125). Kyohokai was founded at the request of Toyota's suppliers (referred to asco-operating factories or kyoryoku kojo) in order to ensure that they could survive thisperiod of hardship. A central task of the Kyohokai, not surprisingly, was to channel rawmaterials and funds, which only Toyota Motors could secure through the rationingsystem, to member suppliers.
With the end of World War II, Kyohokai's central task shifted towards makingimprovements in the technological and managerial capabilities of member firms. Theoriginal members, who formed the Tokai (region) Kyohokai in 1947, were joined bynewly formed Tokyo Kyohokai and Kansai Kyohokai members. The latter memberstended to be larger specialist component suppliers, which developed independentlyof Toyota and were considered more capable managerially. The new group of sup-pliers aroused eagerness among the original members to strengthen their managementsystem.
The first opportunity to make improvements collectively presented itself in 1953,when the prefectural authorities offered a free factory benchmarking service (kojo
Suppliers' associations 661
shindan) to Kyohokai members (Kyohokai, 1967, p. 24^ Wada> 1991). This service waspart of the post-war government policy of rationalising and modernising small andmedium-sized enterprises (SMEs). Because of the large numbers, the Japanese SMEAgency chose the keiretsu group (and the industrial district) as units of diagnosis, thusendorsing the existence of suppliers' associations. The public consultancy offeredconcrete solutions to establishing managerial objectives and production plans, and toimproving productivity and quality at each of the supplier firms. Kyohokai members'effort in implementing the solutions bore fruit in the form of the launch of ToyopetCrown in 1955.
Government policy also encouraged SMEs to form themselves into groups in a moreexplicit manner, through the SME Cooperative Association Law (Chusho Kigyoto KyodoKumiaiHo) of 1949. Co-operatives, once approved and registered publicly, can take outinvestment loans from government financial institutions and receive government sub-sidies for a variety of activities, such as employee training and joint research, in recentyears. Around 47,000 formally registered co-operatives existed in Japan in 1991 (SMEA,1992, p. 106). They are regionally based associations and do not normally have a focusaround a common customer among their members. Co-operatives in the car industry areexceptions to this convention. The 1950s saw the formation of a number of co-operativesin the car industry, by local suppliers to Mazda, Daihatsu, Hino and Nissan Diesel. Atthese assemblers, co-operatives continue to exist, membership overlapping with thesuppliers' associations that were subsequently founded.
A major co-operative association in the car industry that is in effect a suppliers'association at the same time is the Suzuki Motors Cooperative Association. It wasfounded in 1957 when the Shizuoka prefectural office approached Suzuki Motors withdetails of recently enacted laws concerning SMEs. Suzuki cashed in on the financialfacilities offered by these laws to create its suppliers' association in the form of a co-operative. The Cooperative Law requires that the co-operative membership be restrictedto SMEs. However, except for this requirement, the reasoning behind setting up theSuzuki Cooperative was rather similar to that for founding suppliers' associations ingeneral. A document 'The intent to establish the Suzuki Cooperative Association' states:
With a rapid progress in society, every company is facing increasingly tough competition overimprovements in product performance and the expansion of production. Our cooperating factoriesmust adapt to both the economic and technologica] aspects of this situation. As one measure, apolicy shall be hereby implemented, which will promote the welfare of every cooperating factorythrough friendship and mutual help. We intend to promote a higher level of cooperation, and toachieve co-prosperity with Suzuki Motor Company as our parent factory.
Around the same time in 1958, Nissan's plant level supplier associations at theYokohama and Yoshiwara factories, founded a few years earlier, were consolidated intoTakarakai. Takarakai was essentially a gathering of small and medium-sized firms, whosetechnological and managerial capabilities Nissan deemed it necessary to strengthen. Theconcerted effort to improve industrial engineering and to adopt quality control methodsculminated in Nissan's receipt of the Deming Award, the very first in the Japanese carindustry, in June 1960. This spurred Toyota suppliers to do better. From 1961 ToyotaMotors started to take a greater lead in Kyohokai activities, centred around the diffusionof Total Quality Control and Valve Engineering to the top management of core suppliers(Wada, 1984, p. 88). Domestic rivalry was thus a significant factor in intensifying theeffort poured into association activities.
662 M. Sako
However, the 1960s presented an added challenge, namely the liberalisation ofinternational trade and capital markets. In anticipation of open trade, Japanese as-semblers made a concerted effort to improve quality and cost efficiency. There waswidespread fear that if nothing was done, dismantling the protection would severelyundermine the domestic automotive industry. Some suppliers' associations,' such asNissan's Shohokai and Isuzu Kyowakai, were formed specifically to meet this challengeof internationalisation.
To summarise, suppliers' associations spread in the Japanese car industry in threewaves. The first was the years leading up to World War II, when the assembler andsuppliers attempted to forge organisational solidarity to cope with the war effort;materials shortages posed a problem to be overcome. The second wave was in the 1950s,when co-operatives as well as suppliers' associations were formed; co-operatives were thechannel for low-interest loans for equipment modernisation to members, as well as forfinancial help in setting up common services for members. The third wave in the 1960swas associated with the prospect of the liberalisation of capital markets and internationaltrade, and the perceived need to make a leap in international competitiveness. Theassociations in the post war period were mainly private sector initiatives, but favourablegovernment policies towards SMEs facilitated them.
The pattern established in the 1960s was essentially maintained in the 1970s and1980s. In fact, the origins of the associations formed in the last two decades may betraced back to an earlier period. For example, the Mitsubishi Motor Kashiwakai,founded in 1971, soon after the automotive division was hived off from Mitsubishi HeavyIndustries, is a consolidation of factory-level associations, the oldest of which dates from1950.
2.2 Suppliers' association activities in the 1990sToday, every association has its own rules and regulations that spell out the bureaucraticmachinery that supports the association activities. In the case of Toyota's TokaiKyohokai, an Administrative Board, which meets every two months, is in charge ofdeciding the basic policy of the association and of drafting a plan of activities for theforthcoming year. Such a plan is approved at the Annual General Meeting (AGM),which is attended by top managers of all member companies. They listen intently to thekeynote speech delivered by Toyota's chief executive for clues to Toyota's strategicthinking and future direction. At least a dozen more Toyota managers also attend thisAGM. After an award ceremony for best quality or VA. VE suggestions for membersuppliers, the day closes with a dinner party. More detailed information on Toyota'spurchasing policy, production schedules and sales trends is communicated to membersuppliers via the Discussion Meetings (eight times a year) and an occasional lecture ortwo by Toyota managers.
The activity plan is implemented by sectoral groups and functional committees. Mostof the assembler's associations have this dual structure, although the number of groupsand committees vary from association to association. Tokai Kyohokai has three sectorsgrouped according to the type of parts that members produce. Each group meets once amonth, typically to visit a member supplier's factory and to learn from the member'spresentation about problems and achievements at his company the following month.Tokai Kyohokai has three functional committees, on quality, cost, and health and safety.Each meets at least once every two months to study ways of improving the performanceof member companies. The committees may conduct a questionnaire survey of
Suppliers' associations 663
members, and write reports and manuals on best practice examples based on members'own experience. The cost committee also organises VE seminars, and the health andsafety committee runs technical and safety courses for employees of member companies.Lastly, there are monthly newsletters, and baseball and golf tournaments for members.
The above organisational structure is well suited to dealing with the staple issues ofquality, cost, and health and safety. In fact, although slightly different emphases may beput on various aspects of an issue, the themes typically chosen by the suppliers'associations tend to be focused around improving member companies' shopfloorefficiency. This bias is partly because both assemblers and suppliers have felt thus far thatgains can be made most by concentrating their efforts on incremental process innovationrather than product innovation. This was dictated by the riming of the Japaneseassemblers' entry into the car industry (Wada, 1984, p. 97). Another reason whyassociation activities remain focused around process technology is the need to findthemes that are of wide and common interest to the members. Recently, some suppliers'associations, including those at Toyota and Daihatsu, held meetings to discuss how toreduce parts variety. Suppliers may make good suggestions. However, the spirit of openand frank information exchange may be undermined if there is too much of a shifttowards design and development issues that touch on confidentiality in bilateral trading.
2.3. What's in it for assemblers? A trade-off between control and autonomyThe historical overview above noted why suppliers' associations were started. Do thereasons for their founding also explain why they persist? What are the incentives forassemblers to continue sponsoring their association? Are there disadvantages, as well asadvantages, in having an association from the assembler's point of view? Answers to thesequestions differ again from company to company within the Japanese car industry.
At one extreme, Honda, as mentioned earlier, does not have any association eventoday. Honda's late entry into the four-wheel business made it possible to free-ride onother manufacturers' efforts at improving their suppliers. However, late entry does notappear to be a sufficient explanation for the absence of an association, as some assemblershave established their own associations subsequently, Mazda in 1981 being the mostrecent case.
Honda continues not to have its suppliers' association because of a differentpurchasing philosophy that puts much emphasis on competition and equality of trading.To Honda, a suppliers' association seems too old-fashioned and imbued with anundesirable image of top-down control of suppliers. According to a Honda purchasingmanager, an association is not necessary as long as there is close bilateral exchange ofinformation (Nikkei Sangyo Shinbun, 11 June 1991). Besides, it has been argued byHonda managers that, since a suppliers' association is voluntary, Honda cannot initiatesomething that suppliers thus far have not demanded.
The other assemblers that have an association continue to retain much control over itsactivities. For instance, first, the association is typically financed mostly by membershipsubscriptions but is not fully autonomous as the assembler provides a small subsidy outof the purchasing department's budget. Second, the association's secretariat office mayexist in a separate building from the assembler's purchasing department, as at Toyota,but the secretary-general is typically a retiree from the assembler company. Third,although less common today, many of the associations required the assembler'srecommendation as well as the approval of the association's Administrative Board tobecome members.
664 M. Sako
Association activities remain focused around the assembler whose presence is felt at allthe association meetings; no formal meeting takes place without the presence of at leastone representative from the assembler. For example, although the purchasing depart-ment remains the point of contact for the suppliers' association, functional committeemeetings are attended by the assembler's personnel in the relevant functional department(e.g. the quality committee is attended by the assembler's quality control manager).Precisely what the role these managers play has a different nuance from assembler toassembler. Mazda and Nissan, both with recent reorganisations in their suppliers'association, define their managers' role at these meetings as that of co-ordinator, whomay steer discussions and be instrumental in setting themes. Ultimately, the chosenthemes and activities must be useful to the assembler. By contrast, the PurchasingDepartment of Toyota conveyed their wish to withdraw into a supporting and advisoryrole, as member suppliers are considered capable of taking greater initiatives now than inthe past.
Whether leading or advisory, providing support for the suppliers' association isexpensive in managerial time, judging from the frequency of meetings. Weighed againstthis cost are the frequently cited benefits (a) of being able to communicate aboutproduction plans en masse, and (b) of soliciting good suggestions for common problems,such as parts standardisation and pollution control. The lateral communication andmutual learning among suppliers through association activities were also considered tohave benefits that outweighed the potential danger of cartelisation or bargaining bygroups of suppliers.
3. Assessing the role of suppliers' associations: survey evidence
Having examined the historical and contemporary functions of the suppliers' associationfrom the assembler's viewpoint, this section turns to the analysis of the reasons whyJapanese suppliers participate in association activities. In particular, the key questionsare:
(1) What do suppliers regard as the most important benefits of belonging to asuppliers' association? In particular, how important is lateral intersupplier learningcompared with bilateral assembler-supplier linkages?
(2) How do the benefits of belonging to suppliers' association vary according to theassembler and supplier characteristics?
(3) What are the performance outcomes of association membership?
The rest of this paper reports on suppliers' views about the above according to the resultsof a large-scale survey.
3.1 Survey data descriptionIn July 1993 the author conducted a postal survey of around 1500 automotive partssuppliers in Japan with the sponsorship of Mil ' s International Motor Vehicle Program.The sample was drawn from a comprehensive list of all first-tier suppliers of componentsfor cars and trucks in Japan (Auto Trade Journal and JAPIA, 1993). Interviews at thepurchasing departments of all 11 assemblers and at a few first-tier suppliers precededthe piloting of the questionnaire in Japanese. Japan Auto Parts Industry Association(JAPIA) provided feedback and a covering letter to accompany the questionnaire forits members. Respondents (mostly sales and marketing directors) were asked to answer
Suppliers' associations 665
Table 5. Descriptive statistics of association members in the survey
Members Non-members
Sample (% of total) 396(83-9%) 76(161%)Ownership
Percentage of suppliers who are wholly or partially 34-5 29-3owned by the assembler
SupplyAverage number of customers per suppliers 4-1 3-8Average number of other firms supplying the same 2-3 1-8
product to the same customer*Length of trading
Average number of years of trading to date*** 20 years 12 yearsCustomer commitment in future (in years)** 27 years 16 years
Difference between members and non-members is significant according to the r-test at *, 10% level;**, 5% level; ***, 1% level.
the survey for a typical product provided to their most important customer; they hada wealth of experience, having worked an average of 22 years at the company. Fourhundred and seventy-three usable responses were received, constituting a 30% responserate.
As shown in Table 5, 83-9% of the survey respondents were members of theassociation of the assembler for which the survey was answered. Since the survey did notask about multiple membership, 'non-members' may be members of other associations.However, to the extent that respondents were asked to answer the survey for their mostimportant customer, the member/non-member distinction should capture a majordifference between first-tier suppliers. Among the non-member respondents, 44% weresuppliers to Honda, 27% were suppliers to Toyota and the rest were suppliers to otherassemblers.
In this sample, members were not significantly more likely than non-members to beowned by a car assembler, nor was the number of customers per supplier significantlydifferent between members and non-members. Moreover, the number of other firmssupplying the same part to the same customer was actually greater—2-3 companies—formembers than for non-members—1-8 companies.
The most significant difference between members and non-members lay in the recordof long-term trading with their customer to date: over 50% of the members have tradedwith the customer for 20 years or longer, compared with 38% for non-members.Moreover, members' perception of commitment by the customer into the future wasmuch more long-term than was non-members'. This tallies widi the earlier evidence inSection 1 that the membership turnover of suppliers' associations in generally quite low.
3.2. Benefits of membershipThe majority of the survey respondents are association members. However, there isevidence of a significant variation in the incentives for belonging to a suppliers'association. The questionnaire asked respondents to rank the five most importantbenefits of belonging to an association. The options given were the five listed in Table 6,derived from semi-structured interviews with both assemblers and suppliers and the pilottesting of the survey.
3-52
3-32
2-651 59
3-52
3-21**
2-611-63
3-31**
3-35
2-781-51
3-39*
3-03***
2-90***1-79***
666 M. Sako
Table 6. Benefits of belonging to a suppliers' association, by type of firm
Productdevelopment Design
All joint or capability Decline inmembers supplier-led above average benefits
Obtain better information about the 3-98 408** 414** 3-94customer from the customer morequickly
Learn from other members throughexchange of technical information
Receive useful technical assistance fromthe customer through the association
Receive stable orders from the customerMonitor whether the customer behaves
and acts fairly
'We fear that our know-how would leak 2-49 2-50 2-60 2-77***to other members through associationactivities'
'Benefits of association activities have 2-42 2-44 2-24*** n.a.declined recently. We are not veryenthusiastic'
'Association is no longer central in our 2-28 2-29 2-16* n.a.dealings with this customer5
Mean scores of the top five statements in order of importance (5 = most important; 4 = second mostimportant; 3 = third most important; 2=fourth most important; 1= least important). Mean score for thebottom three statements of (5=strongly agree; 4 = agree; 3=neither agree or disagree; 2 = disagree;1 = strongly disagree).
The difference in means between the binary categories in each column (not shown in the table, are column2: average or below-average design capability, column 3: customer-led product development; column 4: nodecline in benefits of the association) is significant at *, 10%; **, 5%; ***, 1%, according to i-test. Pearson'schi-square tests were also performed on the 1—5 scale, which gave similar results of significance.
Dealing with overall averages first, the most popular benefit of belonging to anassociation was better informational access to the customer. Next, 'learning from othermembers through exchange of technical information'—lateral intersupplier learning—was considered the second most important benefit, and more important than the receiptof technical assistance from the customer. Many members apparently regard fellowmember suppliers as a more important source of technical know-how than theircustomer. This lends support to what Fruin and Nishiguchi (1993) call a 'networkmodel' or a 'learning model', in contrast to the bilateral or dualistic model. These topthree reasons were followed by the benefit of receiving stable orders from the customer,and the benefit of being able to monitor the customer's behaviour. The rank ordering ofthese five reasons remain the same for various sub-samples considered below, but thefollowing distinctions may be noted.
Suppliers were asked in the survey about the process of product development. Thosesuppliers who jointly developed with their customer or took entire responsibility fordesign tended to value the first reason—access to information about the customer—morethan did suppliers whose products were developed by their customer. Thus, the type ofcustomer information that is valued by members appears to be about design and
Suppliers' associations 667
development for the next model, rather than, for instance, about the certainty ofproduction scheduling. Suppliers were also asked to assess their own technologicalcapability. Those which considered their product design engineering capability to beabove average valued access to customer information more than did those with averageor below-average capability.
At the same time, it was the below-average design capability suppliers that cited morefrequently the second most popular benefit—mutual intersupplier learning. Moreover,the third benefit, namely the opportunity to receive technical assistance from thecustomer, was more of a concern for suppliers whose products were developed by theircustomers than for suppliers that developed on their own or jointly with their customer.
These findings might lead one to expect that suppliers' associations continued to be ofgreater utility for members with below-average technical capability than for those withsuperior skills. However, that is not the case. In fact, suppliers with superior designcapability disagreed more than did suppliers with average or inferior skills with both thestatements, 'Benefits of association activities have declined recently' and 'Association isno longer central in our dealings with this customer'. This might be because technicallysuperior suppliers are engaged in informal know-how trading among themselves (vonHippel, 1987), while suppliers who have little to offer to others in terms of technicalknow-how are not getting much useful information from other suppliers.
Table 6 shows a clear distinction in the magnitude of benefits between members whofeel that the suppliers' association is of declining use and those who do not. Comparedwith those who saw no decline in the usefulness of suppliers' association, the suppliersthat agreed with the statement 'Benefits of association activities have declined recently'value both technical assistance and intersupplier learning less, fear leakage of know-howto other members more and value the receipt of stable orders from the customer more.Moreover, the function of the suppliers' association as a forum for monitoring whethertheir customer behaves and acts fairly was more highly rated by suppliers experiencing adecline in the usefulness of the association.
Lastly, the magnitude of benefits differed according to which assembler's associationthe respondents belonged to. As shown in Table 7, members of the Toyota associationtended to regard the association as still central in their dealings with Toyota, not leastbecause Toyota members perceived greater benefits in technical assistance from thecustomer than members of other assemblers' associations. It bears out the impressionthat Toyota's association, with its long-standing history and its concerted effort to diffusethe Toyota Production System, is different both in the magnitude of benefits it brings toits members and the function it fulfils.
3.3. Performance outcomes of association membershipThe paper began by posing three (not necessarily mutually exclusive) points of viewconcerning the function of the suppliers' association. The first was an economic theoryfocus on bilateral contracts that renders supplier associations irrelevant (the BilateralContracting Case). A second view was that associations had been useful for improvingsupplier efficiency in the past but not any longer today (the Declining Utility Case). Athird perspective regarded suppliers' association as like an exclusionary keiretsu or a cartel(the Cartel Case).
We should expect the relative economic performance of association members andnon-members to be different in each of the three perspectives. In the Declining UtilityCase, we should expect members to be no different from, or even performing worse than,
3-59
3-54
2-581-42
2-62
2-12
3-43
306**
2-801-59
2-56
2-65***
3-59
3-06***
302*1-71*
2-39
2-41**
3-50
3-28
2-681 59
2-48
2-49'
668 M. Sako
Table 7. Benefits of belonging to a supplier*' association, by affiliation
Toyota Nissan Mitsubishi Allmembers members members (except Toyota)
Obtain better information about the 400 4-23 3-95 3-99customer from the customer morequickly
Leam from other members throughexchange of technical information
Receive useful technical assistance fromthe customer through the association
Receive stable orders from the customerMonitor whether the customer behaves
and acts fairly
'We fear that our know-how would leakto other members through associationactivities'
'Benefits of association activities havedeclined recently. We are not veryenthusiastic'
'Association is no longer central in our 1-96 2-60*** 2-20* 2-37**dealings with this customer'
Mean scores of the top five statements in order of importance (5=most important; 4 = second mostimportant; 3=third most important; 2 = fourth most important; 1= least important). Mean score for thebottom three statements of (5 = strongly agree; 4=agree; 3=neither agree nor disagree; 2 = disagree;1 = strongly disagree).
The difference in means between Toyota members and Nissan members (or Mitsubishi members or allexcept Toyota members) is significant at *, 10%; **, 5%; ***, 1%, according to r-test.
non-members. In the Cartel Case, we should expect members to be reaping monopolyprofit, and possibly better performing in other respects, compared with non-members.The Bilateral Contracting Case remains silent on this issue.
As shown in Table 8, the overall picture of the link between membership andperformance is mixed. First, pre-tax profitability was lower for members than fornon-members. This profitability result refutes a claim that supplier associations might beacting like a cartel reaping monopoly profit. Second, members saw a faster growth ofsales to the customer than did non-members. Third, R&D as a percentage of sales waslower for members than for non-members. This may be because there are 'networkexternalities' among members who can achieve high sales growth without R&Dspending. This may also indicate that there is a group of suppliers who do not wish tobelong to an association as they would do better by appropriating benefits from intenseR&D without fear of leakage to competitors. Fourth, members were not significantlybetter at achieving cost reduction nor in reducing production and delivery batch sizes,according to the survey data.
Does performance vary with different incentives for belonging to an association? Theonly significant distinction was between members who agreed and those who disagreedwith the statement that benefits of the association declined recently. Those perceiving adecline in the usefulness of the association experienced slower growth in both sales to thecustomer and market share, while their costs rose more on average during 1988-1992.
2-6%1-7%17%
7-8%0-8pp
- 0-7%
47 hours14 hours
3-5%2-6%15%
4-1%l l p p
- 0-4%
40 hours18 hours
Suppliers' associations 669
Table 8. Performance outcomes of association membership: survey evidence
Members Non-members
Pre-tax profits as percent of sales ***R&D as percent of sales***Percent of sales in new products since four years agoPercent change in sales to the customer 1988-1992**Percentage point (pp) change in the Japanese
market share 1988-1992Average annual change in costs 1991-1992'Just-in Time' (JIT) production and delivery
Production batch size 1993Delivery batch size 1993
(average number of hours the batch lasts the customer)
Percentage point change measures the difference between one rate and another. For example, the changefrom a 4% margin to a 6% margin is 2 percentage points.
Difference between members and non-members is significant according to the t-test at: *, 10% level;**, 5% level; ***, 1% level.
Conclusions
This paper has presented an analysis of the structure and the functions of suppliers'associations in the Japanese car industry. The underlying set of questions guiding theanalysis was: (a) why did the suppliers' associations start; (b) why do they persist; and (c)what have been their effects?
The major empirical contribution of the research reported here is in its reliance on theresult of a large-scale survey of nearly 500 parts suppliers to all the 11 Japaneseassemblers. These data have made possible the analysis not only of the structure of theassociation but also of the suppliers' perspectives on the role of these associations. Thediversity in their function from association to association is borne out by survey evidencethat complements the case studies carried out on Toyota (Wada, 1984) and onMitsubishi (Smitka, 1991).
The survey provided evidence that the suppliers' association continues to exist in Japannot merely out of inertia but because it is serving a useful function in delivering benefitsto both the assembler and member suppliers. Those suppliers who found the associationto be of declining utility were not in the majority, and were less common among membersof Toyota's Kyohokai than among members of other associations. Thus, suppliers'associations are not a historical anachronism.
The survey also shows that suppliers' associations are not cartel-like entities. Acombination of lower profits, lower R&D expenditure but higher sales growth ofmembers compared with non-members calls for an alternative theoretical framework tothe cartel theory. The recent application of transaction cost economics and game theoryto assembler-supplier relationships in the car industry have contributed much to ourunderstanding of the transactional efficiency of bilateral relationships. However, thesurvey finding that lateral communication and learning between suppliers in the sameassociation is just as valued by members as technical assistance from the assembler callsfor a modification to this bilateral contracting focus in economic theory. That is, supplierrelations in the Japanese car industry should be conceptualised not merely as a nexus ofbilateral contracts but as a network of innovators (Freeman, 1991). The latter framework
670 M. Sako
facilitates the examination of the sources of competitiveness of the Japanese car industrynot only in terms of relation-specific investment between an assembler and a supplier(Asanuma, 1989), but also in terms of informal know-how 'trading' (von Hippel, 1987)and other types of network-based learning among suppliers. This calls for a more explicitrecognition and analysis of externalities to bilateral relationships.
This study indicates that, on the whole, the suppliers' association contributes towardsenhancing the informational efficiency of the industry-wide network. The core nodes inthis network are not so much the assemblers as the twenty or so major componentsuppliers which serve several assemblers at once. The suppliers' associations withoverlapping membership have helped diffuse innovative practices relatively rapidlywithin the Japanese automotive sector. This network for technology diffusion appears tobe as important as the close bilateral assembler-supplier relationship in accounting forthe overall performance of die Japanese automotive industry.
Bibliography
Andersen Consulting 1993. The Lean Enterprise Benchmarking Project, London, Arthur Andersen &Co
Aoki, M. 1988. Information, Incentives, and Bargaining in the Japanese Economy, New York,Cambridge University Press
Asanuma, B. 1989. Manufacturer-supplier relationships in Japan and the concept of relation-specific skill, Journal of the Japanese and International Economies, vol. 3, 1-30
Auto Trade Journal and JAPIA. 1993. Mihon no Jidosha Buhin Kogyo 1992193 (The JapaneseAutomotive Parts Industry), Tokyo, Auto Trade Journal
Cusumano, M. A. 1985. The Japanese Automobile Industry, Cambridge, MA, Harvard UniversityPress
Dodwell Marketing Consultants. 1986. The Structure of the Japanese Auto Pans Industry, Tokyo,Dodwell
Freeman, C. 1991. Networks of innovators: a synthesis of research issues, Research Policy, vol. 20,499-514
Friedman,D. 1988. The Misunderstood Miracle: Industrial Development and Political Change in Japan,Ithaca, NY, Cornell University Press
Fruin, W. M. and Nishiguchi, T. 1993. Supplying the Toyota Production System: intercorporateorganisational evolution and supplier subsystems, in Kogut, B. (ed.), Country Competitiveness,New York, Oxford University Press
Gerlach, M. 1992. Alliance Capitalism: the Social Organisation of Japanese Business, Berkeley and LosAngeles, University of California Press
Grossman, S. and Hart, O. 1986. The costs and benefits of ownership: a theory of vertical andlateral integration, Journal of Political Economy, vol. 94, 691-719
Helper, S. and Sako, M. 1995. Supplier relations in Japan and the United States: are theyconverging? Sloan Management Review, vol. 36, no. 3, 77-84
Hines, P. 1994. Creating World Class Suppliers: Unlocking Mutual Competitive Advantage, London,Financial Times and Pitman Publishing
Ikeda, M. 1989. Jidosha buhin sangyo wo osou shitauke saihen no arashi (A storm of subcontract-ing reorganisation raids the auto components industry, Ekonomisuto, 23 May, 46-51
Ikeda, M. 1992. Jidosha kaihatsu ni okeru sabushisutemu no shinkaiten (New trends in thesub-system of automobile development), Shoko Chukin, February, 3-24
Imai, K. 1994. Enterprise groups, in Imai, K. and Komiya, R. (eds), Business Enterprise in Japan,Cambridge, MA MIT Press
Itami, K. etal. 1988. Kyoso to Kakushin—Jidosha sangyo no kigyo seicho {Competition and Revolution:Corporate Growth in the Automotive Industry), Tokyo, Toyo keizai shinposha
JAMA 1988. Nikon Jidosha Sangyoshi (History of the Japanese Auto Industry), Tokyo, JapanAutomobile Manufacturers Association
Suppliers' associations 671
Kyohokai 1967. Kyohokai no Aywm {The History of Kyohodai), Toyota, KyohokaiLamming, R, 1993. Beyond Partnership: Strategies for Innovation and Lean Supply, Prentice HallMIRI (Mitsubishi Industry Research Institute, Inc.) 1987. The Relationship between Japanese Auto
and Pans Makers, Tokyo, MIRIMiwa 1989. Shitauke kankei: jidosha sangyo (Subcontracting relationships in the car industry), in
Imai, K. and Komiya, R. (eds), Nihon no Kigyo (Japanese Enterprises), Tokyo, Tokyo UniversityPress
Nakamura, T. 1986. Shotoa Keizaishi {Economic History of the Shotva Period), Tokyo, Iwanamishoten
Nishiguchi, T. 1994. Strategic Industrial Sourcing: the Japanese Advantage, New York, OxfordUniversity Press
Piore, M. J. and Sabel, C. F. 1984. The Second Industrial Divide, New York, Basic BooksSabel, C. F. 1989. Flexible specialisation and the re-emergence of regional economies, in Hirst, P.
and Zeitlin, J (eds), Reversing Industrial Decline? Industrial Structure and Policy in Britain and herCompetitors, Oxford, Berg
Sako, M. 1992. Prices, Quality and Trust: Inter-firm Relations in Britain and Japan, Cambridge,Cambridge University Press
Shiomi, H. 1985. Seisan rojisutikkusu no kouzo: toyota jidosha no kesu (Production logistics: thecase of Toyota Motors), ch. 3 in Sakamoto, K. (ed.), Gijutsu Kakushin to Kigyo Kozo {TechnicalChange and Corporate Structure), Tokyo, Minerva shobo
SMEA (Small and Medium sized Enterprise Agency, Japan). 1989. Chusho Kigyo Hakusho (SMEWhite Paper), Tokyo, SMEA
SMEA (Small and Medium sized Enterprise Agency, Japan). 1992. Chusho Kigyo Yoran (Summaryof Small and Medium Enterprises), Tokyo, SMEA
Smitka, M. 1991. Competitive Ties: subcontracting in the Japanese Automotive Industry, New York,Columbia University Press
Ueda, H. 1989. Jidosha sangyo no kigyo kaiso kozo (The hierarchical structure of firms in the carindustry), Kikan Keizai Kenkyu, vol. 12, no. 3, 1-30
von Hippel, E. 1987. Cooperation between rivals: informal know-how trading, Research Policy,vol. 16, 291-302
Wada, K. 1984. Jun suichoku togo gata soshiki no keisei—toyota no jirei (The formation of a'quasi-vertically integrated organisation': the case of Toyota), Academia {Economic and Manage-ment Studies), vol. 83, 61-98
Wada, K. 1991. Jidosha sangyo ni okeru kaisoteki kigyokan kankei no keisei: toyota jidosha no jirei(The evolution of hierarchical inter-firm relations in the car industry: the case of ToyotaMotors), Keieishigaku {Business History), vol. 26, no. 2, 1-27
Williamson, O. E. 1985. The Economic Institutions of Capitalism, New York, Basic BooksWomack, J., Jones, D. T. and Roos, D. 1990. The Machine that Changed the World, New York,
