Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
6 The Sources of Competitive Advantage: Cost and Quality Comparisons Almost 30 percent of all new cars sold in the United States in the spring and summer of 1980 were manufactured outside North America. The recent surge in imported products was influenced by a series of special factors that caused a rapid shift in consumer preferences. Yet the level of import penetration has been growing since the late 1960s. It seems that the long-term success of the Japanese and European producers may be the result of name fundamental factors than the recent shift in preferences or the absence of domestic small-car capacity. We noted in Chapter 4 that domestic producers have had little incentive to excel in small-car production and that at least until the early 1970s the foreign producers typically enjoyed a cost advantage. The evidence also suggests that the most successful foreign producers [e.g., Volkswagen (VW), Toyota, Nissan] have combined lower rote with an emphasis on quality. achieved at the low-price end of the market, but the imports sought to add more performance and quality to the product than the low price itself might have suggested. Rather than pass their cost advantage on to the consumer in the form of an even lower price for an average-quality product, the imports used part of their cost advantage to develop a more competitively viable advantage in product performance and quality. The success of the import strategy provides useful insight into the nature of competition and consumer demand in the market for smaller cars. (The analysis may also apply to other segments, but our focus is on the markets in which domestic and foreign products compete.) The trend in import penetration suggests that the market is sensitive to the price-quality package and that product quality is becoming an increasingly important dimension of competition. It also seems clear that production cost and product quality are closely related and that both must be examined in assessing competitive advantage. Comparison of the relative competitive position of domestic and foreign products is ~ nus, initial penetration was 90
91 in essence a comparison of systems of production. The success of the imports has underscored the obvious point that a production system must be judged in terms of the cost and the quality of its output. The twin issues of cost and quality are central to the future of the domestic auto industry. An assessment of the continued viability of domestic production requires an evaluation of current competitive positions and some analysis of trends in the under- lying determinants. _ . . . . . .. . We first consider the costs of production. our approach Is to draw on publicly available information as well as on data from industry sources to arrive at estimates of the cost of producing a comparable vehicle in both the United States and Japan. We then examine evidence on product quality from a variety of sources. Several dimensions of quality are identified, and an attempt is made to assess the relative position of domestic and foreign products on each. The chapter concludes with an assessment of the sources of U.S.-Japanese differences. It should be noted that we do not attempt to assess the com- parative advantage of U.S. producers in the sense of classical economic trade theory, which depends on relative .. . .. costs of production at home and abroad (e.g., ratio of costs in autos to costs in other goods, versus the same ratio in other countries). Our analysis is focused on the competitive position of U.S. automobile producers relative to their major competitors within the automobile industry. The approach we use here places priority on understanding the characteristics of intraindustry competition, including the role of costs, product quality, and technological innovations. While an analysis of domestic relative costs (i.e., autos versus others in the United States) and relative costs in Japan or Europe would be a potentially useful element in an assessment of trade patterns, such an analysis in the context of differentiated products, economies of scale, and oligopolistic markets is likely to be complicated and is beyond the scope of this study. COMPARATIVE COSTS OF PRODUCTION Over the last several years, information on foreign and domestic productivity and factor prices has been developed that implies a slight advantage for Japanese products and a disadvantage for producers in West Germany and the United Kingdom. These com- parisons are for vehicles available for sale in the United States and thus include the costs of ocean freight and applicable tariffs for imported products. One of the most careful studies of relative costs was conducted
92 by Eric Toder and his colleagues at Charles River Associates.2 Using data for 1974, Toder found a Japanese disadvantage of 3 percent. If Toder's analysis were revised to reflect more realistic transportation costs, the data would imply a Japanese advantage of about 7 percent. In 1978, additional U.S.-Japanese cost comparisons were pub- lished by Ford in a white paper entitled State of the U.S. Automo- tive Industry. Ford estimated the net Japanese cost advantage per vehicle to be $525 on a subcompact-size car landed in the United States. Although methods and sources were not identified, higher U.S. costs were largely due to higher wage rates, a result generally consistent with the analysis of Toder. The difference in the two analyses could be explained by differences in time period, intervening inflation, and changes in productivity. The notion that the landed-cost differential between U.S. and Japanese products is $500-$600 recently found its way into con- gressional testimony. Speaking before the Subcommittee on Trade of the House Ways and Means Committee. Abraham Kelp _ _ ~ ~ , ~ ~ ~ _ ~ ~ ~ ~ . . ~ Assistant secretary ot Commerce for International Economic Policy, summarized what appears to be a concensus view: Average hourly compensation (including fringe benefits) in the Japanese auto industry in 1979 was $6.85--half of the 313.72 hourly compensation in the U.S. auto industry. Present indications are that productivity in the U.S. and Japanese auto industries may be roughly equal. On this basis Japanese producers appear to have had an $860 labor cost advantage per car in 1979. Other differential costs (principally the higher cost of steel in the United States) may have added $100 per car to the U.S. cost. As freight and insurance on Japanese cars averages ii400, the apparent cost advantage to Japanese producers may have been $560 per car in 1979. The actual advantage may have been considerably less, for the above calculations do not take into account energy costs, capital costs, and the costs of other production factors--some of which are cheaper in the United States than in Japan.3 It is our view that the estimates presented by Katz and his suggestion that actual differences might be even lower constitute an understatement of the current cost advantage of the Japanese. Not Shiv clip the Him;c fail try rmfI^~1 Ran rip ^c of ^rr~r~~r~_ ~ ~ ~ _ ~ & ~ _ 1 ~ ~ ~ ~ ~ V ~ a_ V ! ~ · !J ~ ! 1 . . . . . . . sallon nut they also tail to capture important differences in production processes that result in higher productivity in Japan. Estimates that reflect these differences have been developed using a variety of methods. We have estimated the productivity and cost differential using both a macro, economy-wide approach
93 and a "bottom up" approach with microdata. We have taken an industry-wide perspective, using publicly available sources, and we have analyzed annual reports. Comparisons of this sort involve several difficulties. The automobile manufacturers of the United States and Japan produce a different mix of products and have organized production in different ways, particularly in terms of vertical integration. - ~ Productivity comparisons are also signet ~- cantly affected by differences in capacity utilization that have been substantial in recent years. While attempts have been made to correct for these factors, even the most careful comparison requires judgements and assumptions that affect the results. A detailed description of the various kinds of analysis used is presented in Appendix A. Evidence from Alternative Perspectives Table 6.1 outlines the various perspectives taken and summarizes the basic results. The analysis suggests that the Japanese enjoy a landed-cost advantage of between $700 and $1500 per small vehicle. These estimates are larger than those used in congres- sional hearings during 1980. Furthermore, the immediate sources of the Japanese advantage may be quite different. Those analyses that focused on the auto sector found sizeable differences in labor hours per vehicle (a measure of productivity) along with differ- ences in employee costs and other prices. The industry level analysis (which includes suppliers) found a 20-25 percent Japanese advantage in productivity, while examination of specific plants and processes revealed an even larger productivity gap. The wide disparity in estimates of the Japanese productivity advantage underscores the difficulty of making precise calcula- tions, particularly in this context, where the industry structure is different and hard data are relatively scarce. Indeed, precise order of magnitude and the confidence which industry panel members place in the estimated cost difference (i.e., $1200-$1500) comes much more from internal studies, using confidential and proprietary data, than from the relatively rough analysis presented in Appendix A. Some of the calculations in Appendix A are consistent with the internal evidence, and all of them point to a sizeable U.S. cost disadvantage that appears to reflect differ- ences in both prices and productivity.4 ASPECTS OF PRODUCT QUALITY When VW first made significant penetration into the U.S. market, its strategy established a formula for success that has been refined
94 m .' ~ ° E ' ~ ° to E C o o ·~> =` a) . ~ .= 3 E B. o z) o ~ 3 . o Ct ~ .~.= g ~ ~ _ ._ ~ ~ lo, ~ · . ~ ~ o ~ ~ o Cal ~ ~ ~ ~ 3 o ~ o ~ c Cal - Cal U. o Ct Ct C~ C~ ~ ~ ~ o =~ E E 2 ~ E 8 E ° Ct . E E c, ~ ~ S !~ 5 1 ~ ~ ~ E <,, E E E a E E ~ E ~ ~ 0 ~ ~ ~ ~ _ ~O~ ~ I_ O ~ r 3 ~ .S . ~ Ct ~ o C~ C~ ;> Ct o o ~ _1 & C~ o ~o C:S ;> Ct ~ o (L) O C~ C~ ~ ~ . - o ~ ,, o . o o o ~ ~ ~ ~ o ~ ~ ~ a.> ~ .= . b.> ~o ;~ o ~ o a~ ~ o ~ o C.) ° ~¢ U. ~ $ ~ ~i o U. ;s C~ o x . ~L a, v, . . o u,
95 and extended by the Japanese. A critical element in that strategy was the production of a vehicle that the market perceived to be of high quality. Beginning with the VW Beetle in the late 1950s~ the . word used most often to describe the character of imported products has been "workmanship," which connotes attention to detail and care in production and quality appearance. The view is now widespread that quality defined in these terms has been a significant factor in the recent success of the Japanese. It seems clear from recent statements of industry executives that an improvement in quality will be an important aspect of any improvement in the U.S. competitive position. In this chapter we identify dimensions of quality that appear to be significant and present evidence about the relative U.S. position. 1 ~ Definitions Any attempt to define and evaluate the quality of a complex product such as the automobile must deal with the supposed distinction between perception and reality. It has been argued that the Japanese and European advantage in quality is not "real" in an objective sense but is only a perceived advantage. The implication is that through advertising and other forms of public persuasion the importers have created an image of quality that colors consumer perception. This argument misses the point. In the marketplace, perception is reality. The competitively impor- tant dimensions of the product are not those established by experts, nor are the key differences between manufacturers those determined by an objective evaluation. Competitive advantage accrues to those whose products are perceived by the buyers to be of higher quality. v hi, Seen in these terms, quality is simply whatever the market defines it to be. A manufacturer can go to great lengths to offer a car with clearly superior rustproofing, but if corrosion protec- tion is not an element of the market's definition of quality, little competitive advantage will be obtained. There appear to be three dimensions of quality that the market regards as significant. We propose to define each dimension and present some evidence on the relative U.S. position. Assembly Quality This category harks back to the notion of "workmanship" identi- fied earlier. It has been described as the "fits and finishes" dimension and includes such things as body finish, squeaks and
96 r attles, the alignment of doors and hoods, and paint quality. Within the industry, assembly quality is usually defined in terms of "building to spec," that is, making the vehicle as specified in the design. This definition focuses attention on the work performed on the assembly line and on the extent to which components and materials meet specifications. But it is clear that the design and thus the specifications themselves are also a significant factor in determining assembly quality. Not only does the design affect the ability of the manufacturing organization to achieve high-quality assembly, but the standards established in the specifications may not be as exacting as those of competitors. Available information suggests that U.S. producers' assembly quality falls short of the implicit market standard set by the imports. Table 6.2 presents ratings of the condition of selected vehicles at delivery and the number of defects after one month of service. Imports have a clear advantage in both measures, although the gap is largest in defects after one month. These data suggest that consumer perceptions are consistent with actual experience with purchased vehicles. Reliability The automobile is a collection of complex mechanical and electrical systems that are subjected to enormous stress--wide variance in temperatures, short bursts of heavy use followed by long periods of inaction, and so forth. To be even minimally competitive, cars must achieve a very high degree of reliability-- that is, the ability to function as designed on demand. In these terms, reliability applies both to individual components and to entire vehicle systems. Failure to function as designed makes the , vehicle less useful (at times unuseable), and repairing a malfunc- tion is often a time-consuming hassle. Reliability is thus a critical dimension of overall quality. To measure reliability we have used the repair incidence data published by Consumer's Union. These data were not drawn from a random sample of all owners but rather from the subscribers to Consumer Reports. They may not be representative of experience generally. ~l~he basic data cover repair frequency of mechanical systems and components and the body (structure, finish). Ratings are given in five categories average, below average, far below average, above average, and far above average. Beginning with a score of O for far below average, we have assigned values of 5, 10, 15, and 20 to the other categories. A total score for each vehicle was obtained by summing the scores on individual body and mechanical systems. The results are reported in Table 6.3 for selected makes.
97 It is apparent that imported products have achieved repair records that exceed those of the domestic manufacturers, in some cases by substantial margins. While the imports have an advan- tage in both body and mechanical systems, the superiority of the foreign products is most pronounced in the body category. This is consistent with earlier evidence on assembly quality. In mechani- cal systems, reliability of some of the domestic and imported products is actually quite close. Given the nature of the data and size of the differences, strong conclusions about an overall import advantage in mechanical system reliability does not seem warranted. TABLE 6.2 Rating of Assembly Quality: U.S. Versus the Imports, 1979 Condition of Car at Delivery Condition of Car after One (scale of 1-10; 10 is Month of Service (number of Vehicle Category excellent) defects per vehicle shipped) Aggregates Domestic Imports Subcompact 6.4 7.9 Compact 6.2 7.7 Midsize 6.6 8.1 Standard 6.8 Specific Models Omni 7.4 4.10 Corolla 7.8 0.71b Chevette 7.2 3.00 Pinto 6.5 3.70 Rabbit (U.S.)a 7.8 2.13 Fiesta 7.9 N/A Civic 8.0 1.23C Horizon 7.5 N/A Colt 7.8 N/A a European Rabbit averaged 1.42 defects per vehicle shipped. b Toyota average. c Honda average. SOURCE: Aggregates: Rogers National Research, Buyer Profiles, 1979; Models: Industry Sources.
98 TABLE 6.3 Ratings of Body and Mechanical Repair Frequency, 1979 (10 = average; 20 = best; 0 = worst) Make Body Mechanical Domestic Buick 10 10 Chevrolet 4 Dodge 8 8 Ford 9 7 Lincoln 10 10 Oldsmobile 11 9 Volkswagen 14 11 Imports Datsun 14 11 Honda 16 12 Mazda 18 13 Toyota 17 12 Volkswagen N/A N/A Volvo 1 6 1 1 NOTE: The data cover repair frequency of mechanical systems, components, and body (structure and finish). Ratings are given in five categories: average, below average, far below average, above average, and far above average. Beginriing with a score of 0 for those that are far below average, we have assigned values of 5, 1¢, 15, and 20 to the other categories. A total score for each vehicle was obtained by summing the scores of individual body and mechanical systems. SOURCE: Consumer Reports, April 1979. Durability There is little evidence about the long-term durability of Japanese products and thus little basis for comparison. It does appear that U.S. products have superior corrosion protection and that basic components and systems may be more durable. Customer Loyalty Perhaps the most significant test of quality production an d customer satisfaction is loyalty, the willingness of buyers to purchase the same car again. The data presented in Table 6.4
99 TABLE 6.4 Customer Loyalty: Would Buy Same Make/Model Again (percentage) Domestic Imported Total Subcompact 77.2 91.6 81.2 Compact 74.2 91.4 72.4 Midsize 75.3 94.5 76.9 Standard 81.8 Luxury 86.6 94.6 87.2 Total 78.7 91.8 SOURCE: Rogers National Research, Buyer Profiles, 1979. generally confirm the evidence on assembly quality and reli- ability. The fraction of owners willing to buy the same make again is much higher in the import group. Since the Japanese dominate the import category, these results underscore their f ormidable competitive advantage. Not only are their costs significantly lower, but the quality of their products is higher EXPLANATION A N D PROG NOSIS . It has become almost commonplace to cite the superior quality of Japanese cars as a rationale for their competitive success. With the evidence on productivity and costs, it appears that the Japanese competitive position is buttressed by a significant cost advantage as well as by higher-quality production. On both counts the gap between the United States and Japan is significant. While there is little evidence of a serious attempt to exploit the cost differential through aggressive pricing, it is clear that Japanese manufacturers can absorb very large increases in costs without raising prices and still obtain higher margins than their U.S. competitors.5 And it is equally clear that a sufficient margin exists for even more costly improvements in performance and quality. Japanese quality levels, however, are already perceived to be a cut above domestic production. With their emphasis on quality and performance the major Japanese firms have acquired a kind of "reputation capital" that enhances an already formidable competitive position. Popular accounts of the emergence of Japanese producers as first-rate, worldwide competitors tend to emphasize the impact of new automation technology (e.g., robotics), strong support of the central government (i.e., "Japan, Inc."), and influence of Japanese culture (i.e., a dedicated work force). There is no doubt that these factors have played some role. Yet, it is our view that the
100 sources of the Japanese advantage are not to be found in such factors. Rather, they are rooted in a commitment to manufac- turing excellence and a strategy that uses manufacturing as a competitive weapon. It may seem odd to think of manufacturing as anything other than a competitive weapon. After all, "manufacturing" refers to the production and distribution of the product--essential features of competition. Yet the history of the automobile market in the United States suggests that by the late 1950s manufacturing had become a competitively neutral factor. This is not to imply that it was not important; indeed, the Big 3 expended great resources in improving technology and productivity. The point is that none of the major producers sought to achieve a competitive advantage through superior manufacturing performance. Except perhaps for economies of scale, which are affected by manufacturing aec~slons, the basis of competition was located outside manufacturing--in marketing, styling, and the dealerships. But the Japanese advantage originates precisely in manufac- turing operations. Productivity and quality are determined in the very heart of the operation, in the interaction of Deoole. m aterials, and equipment. It is in the management of these elements that the Japanese have excelled. And it is the dictates of strategy that have provided the impetus for that excellence.6 The strategy of the Japanese producers was first and foremost an entry strategy. The fact that they started from the ground up in the U.S. market influenced their choices in all dimensions of competition. When the Japanese sought to penetrate the U.S. m arket it would have been foolish to try to compete with the domestic firms on their terms. Just as General Motors (G M) avoided head-on competition with Ford in the 1 920s, so the Japanese approach avoided status quo competitive behavior in the 1960s. The domestic market was dominated by the large car, the annual model change, and the "boulevard ride." The new entrants had neither the experience, the production systems, nor the resources to compete on those terms. As with other imports, the J apanese sought out a niche in the small-car segment. Having learned from their early failures (the first attempt at penetration failed on the strength of a low-performance, low-quality product) and the success of VW, both Toyota and Nissan concentrated on establishing a dealer network and on producing a high-quality, solid-performance small car. It was essential that the level of quality and performance be noticeably superior. Otherwise the new product lines were destined to be lost in the competitive shuffle. Moreover, reputation for quality and performance was essential for success over the longer term when entry into higher-margin niches (sports cars, high-performance sedans) was . . envlslonec ,. ~ . . rrid~
101 Explaining the Performance Gap If competitive strategy provides the broad driving force for excellence in manufacturing, what explains observed U.S.- J apanese differences in performance? What aspects of the production process should be singled out for particular notice? To cast some light on these issues, we have identified several char- acteristics of the production process that may be important in explaining differences in productivity. The productivity of an operating system--in this case the n umber of employee hours required per vehicle produced--is determined by the state of technology (both product and process by the amount and quality of inputs; and by the way in which the resources are combined, organized, and managed. At its most basic level the productivity of an existing operation and tech- nology can be improved either by improving the quality of r esource (e.g., hiring more highly skilled workers, using better materials, and so forth) or by more effectively utilizing the existing set. The latter may involve things like changes in supervision, changes in the procedures used to control materials, or a host of other management and organizational factors. Productivity can also be enhanced by introducing advanced technology--new equipment, new products, or new processes and technologies. These basic determinants--technology, resources, and manage- ment systems--can be used to compare and contrast production operations. Our analysis of the U.S.-Japanese productivity gap in auto production is organized around seven factors that have been grouped into three categories: process systems (process yield, quality systems), technology (process automation, product design), and workforce management (absenteeism, job structure, work pace). Any attempt of this sort runs the risk of arbitrary categor- ization. While useful in clarifying determinants, it should be recognized that many of these are closely related. Table 6.5 provides definitions of the factors affecting the productivity differential, along with a brief statement of comparative practice in the United States and Japan. The selection of the factors, their definition, and the comparisons are based on discussions with a panel of industry experts. We also asked the panel to rank the factors in order of their importance; some members of the panel provided a percentage allocation. The rankings are presented in Table 6.6. Perhaps the most striking finding in the panel's assessment is the relative unimportance of the factors connected with tech- nology. Neither automation nor product design is accorded a large measure of explanatory power. Despite the publicity devoted to r obotics and advanced assembly plants, such as N issan's Zama
102 Cal ~4 oo - o Cat o o Cal so Cal .o o sit ho on A · _ U) , o V) ._ Ad. o ~ ._ ~ ~ EM O C) U) . _ c _ - o _ Ct Cd ._ .O , ._ Ct no o c t= Cal U. ~ ant 1 1 ~ ~ |~3 1~5 A. 2 ~ ~ · ~ ~ ]~ i~ I g1~3~{ C ~ o ,C ~ -= .~ ~ ~ .C" ~ ~ .. _ . ~ - _ ~ , F - o E E 2= _ ~ ~ ,_, ~~ ~ ~
103 x ·~> Ct a' U. .~ . . Ct ~ 1 ~ | sol ~ c' m I ~ C V) ~ U. =i E ._ ~ o Cal Cal Ed oo sit os _ _ ~ at so C) ._ ~ ~ 3 CSS - oo 53 _ =: .= C~ 3 ~ C~ oo q.) 3$ o - C) C~ o o ~ ~ o o ~ ~ ~ _ _ _ 1 1 1 1 1 1 1 o o o ~ o o C~ _ o ~ o . . . . . . _ C~ o ~ o ~oo o d" ~ _ _ _ _ _ o o V~ ~ o o o ~ ~, _ _ _ ~ ~ d" ~ t _ ~ C~ o o ~ ~ ~ ~ ~ r~ } o . - c~ cq =~ a ~ a a ~ 0,c, ~D O ~ O O ~ ~ ~ ~ ~ ~ 3 . . . _ ~ r~ . . . . o ~; .= ~ _ ;> U, ~ ·= a, _ ._ · & X a., Ct ~D · aq o ._ <_ _ _ ·~ ~ , o C) ~ Ct o ~
104 facility, U.S. firms appear to have maintained comparable levels of advanced process technique and equipment. The panel's assessment is buttressed by evidence presented in Appendix A that suggests that the Japanese producers may use less capital per vehicle than their U.S. counterparts. While it is true that capital-labor ratios are higher in Japan, the large labor productivity gap cannot be explained by simple capital-labor substitution. T ~ ~ ~ one comparison thus makes clear that an explanation of the productivity gap must be found in the quality of resources and m anagement systems. The panel was unanimous in giving top billing to a factor we have labeled "process yield" but that is really an amalgam of several management practices and systems related to production planning and control. The "yield" category captures Japanese superiority in operating their processes at a high level of good output over extended periods of time. Although engineering (i.e., machine cycles, plant layouts) is of some importance, the key to Japan's lead in this category appears to be the interaction of the material control system, maintenance practices, and employee involvement. Figure 6.1 graphically portrays the determinants of annual output of good parts (from a representative production process) and indicates some of the management practices and systems that lead to superior performance in Japan. The key to the material control system is the concept of "just in time" production.7 Often called "Kanban" (after the production cards or tickets used to trigger production), the system is designed so that materials, parts, and components used at a given step in production are produced or delivered just before they are needed. Thus, stages in the process (including suppliers) are tightly coupled, with very little work in-process inventory. Suppliers must therefore make frequent deliveries of parts, and lot sizes must be small to accom- modate product variety. It is the Japanese view that reduction of decoupling inventory exposes "the real problems"--waste of time and materials, imbalance in operations, defective parts, equip- m ent operating improperly, and so forth. (Table 6.7 provides comparative data on inventory levels. These data show that dramatically less inventory is used by Japanese firms in the production of automobiles. This is true whether one looks at the process as a whole or at specific plants.) With small buffer stocks the production system will simply not work if there are frequent or lengthy breakdowns. Thus, the just-in-time approach exposes opportunities for reducing waste and solving problems, while at the same time creating pressure for maximizing uptime and minimizing defects. Maintenance programs, preventive and scheduled, are therefore pursued vigorously. Plants operate with only two shifts, and equipment is maintained during nonproduction . . _
105 DETERMINANTS Process Yield rated machine speed (total X parts per hour) . 1-defect rate up lime X good par s/ (hours per year: total parts) ] annual output (good parts per year) annual outpu (good parts per year) Japanese U.S.-Japan dif- Maintenance: Materials: Results: Management ferences are · Preventive/ · Supplier linked · Significantly Systems not extensive operator to final assembly higher yields involvement schedulepress for processes/ · Shifts scheduled for zero defects equipment to allow frequent maintenance Materials: Line Stops: · Just in time · Worker control productionmin over qual ity Wl P inventory on I ine · Expose and attack · Expose problems problems FIGURE 6.1 Japanese management systems and the determinants of process yield. time. The result is a much lower rate of machine failure and breakdown. Pressure for defect elimination is reflected in relationships with suppliers and in-work practices on the line. "Just in time" production does not allow for extensive inspection of incoming parts. Suppliers must, therefore, achieve highly demanding quality levels, consistently and reliably. The major Japanese manu- facturers work closely with outside vendors to make sure that responsibility for quality is felt and acted upon at the source of product. This same approach--quality control at the source--is used in production on the line, where workers have the authority to stop the operation if they spot defects or other production problems. Worker-initiated line stops are central to the concept of Jidoka: making problems visable to everyone's eye and stopping the line if trouble occurs all that. meth~ric ;`nH talc to vail stops are Jidoka. O , · , ~ The basic thrust of the Kanban system and the concept of Jidoka are to eliminate waste, expose problems, and conserve resources. This is not simply a different technique of controlling production, but a very different way of managing the production process. It is clear that these systems interact with other factors in our list of productivity determinants. Separating their effects f rom the effects of quality systems and job structure, for example, is somewhat arbitrary. The Kanban-Jidoka system uses
106 TABLE 6.7 Inventory Comparisons United States and Japan LeveVProcess Japan United States 1. Plant and Process Inventories Assembly plant component inventories (equivalent units of production) heaters radiators brace drums bumpers Front-wheel-drive transfer case in process parts storage by operation (number of parts) mill drill ream and chamfer drill mill, washer, test assemble finish Total 2. Company Inventories Work in process inventories per vehicle 1979 1980 Work in process turnsa 1979 1980 1 hour 2 hours 1.5 hours 1 hour 11 13 24 10 6 7 79 $80.2 $74.2 40.0 46.1 5 days 5 days 3 days 240 200 196 205 40 96 87 1064 $536.5 $584.3 12.1 13.4 a Defined as cost of goods sold divided by work in process inventories. SOURCES: 1. Industry sources (data provided by panel members); 2. Annual reports for representative producers. fewer inspectors, and its success requires broader and deeper jobs. Seen in this light, the panel's high ranking of "process yield" and the relatively lower importance attached to job structure and quality systems may reflect the difficulty of separating the three factors and a tendency to ascribe to "process yield" some of the i mpact of the other two. Indeed, it appears that job structure plays an important role in explaining observed productivity differentials. We have already noted two features of the Japanese system (maintenance prac- tices and Jidoka) in which jobs are designed to involve workers in a variety of tasks. The effects of structure, and the differences in management style and practices that go with it (fewer layers of management, more managing from the bottom up), extend to other
107 aspects of production. Quality circles or "small group involve- ment activities" deal with such questions as layout, process methods, and automation. Such involvement appears to be an important factor in obtaining relatively high levels of commit- O ~ , O ~ ment and motivation. The nature of worker-management relations in Japan is further suggested by much lower levels of unexcused absence than that f ound in the United States. The panel accorded absenteeism second billing in its rankings, primarily due to the need to carry redundant workers in order to cover for unexpected absence. Appendix B provides an analysis of this effect. In general, absen- teeism influences costs, not only through redundant labor but also through fringe costs of the absent group as well as indirect effects such as scrap, reduced learning, and so forth. It appears that absenteeism may actually account for as much as 10-12 percent of the cost gap. G iven the impact of absenteeism and the effects of job structure and the workforce influence in "process yield," it is clear that workforce management must be a significant factor in explaining the Japanese cost advantage. Likewise, an attempt to explain quality adherences would certainly accord a major influ- ence to the work force and its management. It seems evident. therefore, that in concert with different systems of production management and control, the work force plays a central role in the Japanese competitive advantage. Much has been made recently of the enormous capital expen- diture programs of the U.S. manufacturers. Indeed, the fact that we have identified the Japanese advantage as a "software" rather than a "hardware" problem in no way implies that new technology could not be effective in raising relative U.S. productivity. Yet it is unlikely that a substantially improved competitive position for U.S. production will be secured only through new bricks and mortar and new machines; comparable processes and machines are avail- able around the world. At least as far as advances in productivity and quality are concerned, new "software" (new approaches to the management of people, materials, and processes) seems essential. The next chapter examines industry experience and prospects in one of these dimensions--workforce management. ~ `, ~ . _ _ _ . . ~ ~ .. ~ . . N OTES . 1. The classical economic theory of international trade posits a world in which trade based on differences in national factor endowments takes place in a competitive world economy. This type of analysis focuses on the long-run equilibrium properties of trade and is essentially static in nature. Recent developments . . . · . .
108 in the theory of international trade have taken a more dynamic perspective and have introduced innovation, product differen- tiation, economies of scale, and oligopolistic competition as important elements in determining the pattern of trade. See the papers by Krugman (1980) and Lancaster (1980~. The more recent work has focused on explaining intraindustry trade and has incor- porated direct foreign investment. Once the assumptions of perfect competition and homogeneous products are abandoned, the notion of comparative advantage becomes more complex. Instead o f simple, static comparisons of relative costs of production, which are determined by national differences in factor endow- ments, the characteristics that determine comparative advantage in the more complex models include differentiation of products, innovative capability, and the nature of domestic competition, in addition to production costs. These characteristics may change over time and may be endogenous. The point is not that relative costs of production are irrelevant, but rather that they are but one element in the determination of a broader notion of comparative advantage and that a full-scale analysis of that broader notion is likely to be complicated. Taken in isolation, relative costs in the foreign country could give misleading indications about trade patterns. Nonetheless, the absence of a comparative cost analysis makes the study less conclusive about the trade pattern than it would be if such analysis had been carried out. Furthermore, a study of comparative costs may put the policy issues in a difference light and would therefore be a useful area for further research. For a recent review of these issues, see Whitman (1981~. 2. See Toder (1978~. 3. See Katz(1980~. 4. Given the differences in estimates of the productivity gap, it is difficult to divide the overall cost difference into a wage portion and a productivity portion. In Appendix A, for example, the industry-level analysis attributes about 20 percent of the cost gap to productivity, while the OEM-1 evel analysis estimates productivity's share to be 38 percent. (These are derived using standard cost-accounting techniques for assigning variances.) In any case, it is clear that both wage and productivity differences are important. 5. There is little evidence that aggressive pricing has been practiced by the Japanese in the U.S. market in Europe; however, the Japanese have priced their products somewhat below comparable domestic vehicles. 6. While the emphasis here is on entry strategies, it should be noted that competition in the Japanese domestic market also had a strong influence on the development of manufacturing capability. 7. For a review of the "just in time" production system, see Monden (1981~.
