Contents

1.  Introduction  3

2.  A brief introduction to industry lifecycle theory  4

2.1  The approach of Gort and Klepper to the industry lifecycle  4

2.2  Hirschman’s alternative approach to industrial development  5

3  Lifecycles in the world automotive industry  8

3.1  Europe  8

3.2  The Americas  10

3.3  Asia  12

4  Learning effects in the world passenger car industry  16

5  Summing up  19

References   20

2

1 Introduction

The economic theory of industry lifecycles is a standard tool to explore the historic development of specific industries. Its basic idea is that an industry develops according to a typical, sequential pattern of emergence, rise, maturing and finally decay. This article compares conventional industry lifecycle theory as represented by the seminal approach of Gort and Klepper (Gort / Klepper 1982) with an alternative explanation of industry evolution, i.e. the application of Hirschman’s theory of economic development (Hirschman 1958; Hirschman 1968) to specific industries. The world passenger car industry and its long-term history serve as the case of empiric reference. The analysis is non-technical and rests on a limited set of macroscopic parameters (number of manufacturers, market entry and exit, output) in the World’s regions (Europe, the Americas, Asia including Oceania). After more than 120 years it is a fair question where the automotive industry stands in its lifecycle - irrespective of the convulsions of the current demand crisis. Under the impression of lifecycle theory one may hypothesize that the automotive industry has already reached its dawn. At least in the Western hemisphere, most markets for light motor vehicles have reached a state of saturation. The multiplicity of manufacturers and technological approaches, which characterized the automotive industry’s pioneer days, has long given way to a small number of giant corporations (Möser 2002: 35ff.).

Products differ mainly because of marketing efforts but not so much because of truly new ideas for vehicle concepts or technology. There are scholars who consider motor manufacturing to be among the lead sectors in the “Age of Mass Production” or the “Fourth Long Wave” of technological change. The “Fourth Long Wave”, however, ended back in the 1990s to give way for the “Age of Microelectronics and Computer Networks” which are supposed to be leading technologies for the 21 ^st century. ¹ Reason enough to hypothesize that the automotive industry has passed its pinnacle.

However, this is not the whole story. There are still emerging automotive markets (e.g. in Asia) where motorization is progressing, where new markets are being created and where new manufacturers are entering business. It might well be that a rejuvenation of the automotive industry is taking place there - a rejuvenation which is closely connected to the emergence of newly industrializing economies.

¹ The first long wave of technology was brought about by the Industrial Revolution with the production of textiles as a lead industry (1780s - 1840s). The second wave appeared with the age of steam power and railways (1840s - 1890s), the third with the age of electricity and steel (1890s - 1940s). The age of mass production with automobiles and synthetic materials as lead industries began in the 1940s (Freeman / Soete 1997: 17 ff.).

3

2 A brief introduction to industry lifecycle theory

Cycle patterns of economic development are quite an old topic in economic theory and statistics. Prominent ideas of macro-cycles or “long waves” which drive secular ups and downs of the World’s commercial activity have been brought forward by Kondratieff (Schumpeter 1954: 1158) (with the automotive industry being part of the fourth Kondratieff which peaked in the 1960s) (Freeman / Soete 1997: 19; Nefiodow 1998: 157). More detailed studies focusing on the historic cycles which specific industries go through have been introduced to economics, e.g. by Burns (Burns 1934) or Fabricant (Fabricant 1940). Kuznets’s seminal work on industrial evolution depicts the paths of specific industries’ historic development by means of logarithmic functions (Kuznets 1967).

In the 1980s, management theory and sociologists transformed the concept of the learning curve into a theory of industry lifecycles by stressing the idea that, in the course of time, agents learn about generally accepted ways of developing, producing and marketing specific goods. A sort of commercial orthodoxy evolves which typically coincides with increasing capital endowments and economies of size. Commercial orthodoxy reflects the progressing maturity of an industry. A prominent example for the application of such approaches to the automotive industry was Abernathy’s analysis of the productivity-dilemma, i.e. the trade-off between increasing capital intensity and a decreasing propensity to innovate (Abernathy 1978). The merit of these approaches has been to include the behavior of economic agents into the hitherto more or less descriptive analysis of the industry lifecycle. Nevertheless, in terms of methodology they are eclectic and mono-causal.

2.1 The approach of Gort and Klepper to the industry lifecycle

In a seminal paper on “Time paths in the diffusion of product innovations” Michael Gort and Steven Klepper brought forward a truly comprehensive economic theory of the industry lifecycle (Gort / Klepper 1982). In a nutshell, this approach assumes that the longer a company remains in business, the more it learns about efficient production and at any point of time older firms face a higher likelihood to further remain in business than younger ones. Gort and Klepper identified five typical stages of the industry lifecycle: (I) A creative entrepreneur enters the stage with an innovative product. If his marketing efforts are successful and the product finds its market, other agents will be ready to imitate the product (II) In the next phase the number of producers steeply increases due to the market entry of imitators

(III) Later, the net entry rate approaches zero. The number of new entrants tends to be more or less equal to the number of companies which already have to go out of business

(IV) New entries are outnumbered by market exits in the next phase. The total number of producers begins to decrease (“industrial shake out”) (V) Finally, the total number of producers stabilizes at a comparatively low level and remains constant as long as market fundamentals do not change As time elapses and an industry moves through this standard scheme stage by stage, the nature of knowledge changes which is relevant for setting up efficient, state-of-the-art processes for manufacturing and product development. During the early phases of new industries, almost all the necessary knowledge may be acquired from outside sources, e.g. from scientific institutions or publications. Despite of being highly sophisticated in many cases, it is transferable and explicit. As products and processes mature in the course of time, the focus shifts towards “know-how” or implicit knowledge which is not easily transferable.

4

Now, successful companies will excel through the fine tuning of efficient processes and intricate marketing strategies. The underlying knowledge may no longer be acquired from outside sources but rather needs to be internally accumulated by (time consuming) experience. The more an industry’s knowledge focus shifts from explicit to implicit knowledge, the more are new entrants in disadvantage compared to more experienced incumbents. The likelihood that new entrants make profits which are sufficient to survive decreases over time and with it decrease incentives for market entry. Consequently, the number of new entries shrinks and more market exits occur until some kind of an equilibrium number of producers is reached. Put into a total-market perspective, the maturing of an industry over its lifecycle has far reaching consequences. Maturing is almost inevitably associated with a loss in competitive diversity and an increase in homogeneity: The number of independent suppliers shrinks and with it decreases the degree of competitive, technical and aesthetical differentiation. Fehl emphasizes that any industry’s economic viability depends on the degree of diversity which it is able to bring about compared to other industries (Fehl 1983; 1986a; 1986b). If diversity falls below a certain threshold, competition within the affected industry will become sluggish and tacit collusion tends to be the dominant mode of behavior. The more sluggish competition becomes, the less is the industry able to attract new capital investment. Incumbents begin to invest their returns in other industries instead of spending on their home turf, e.g. for R&D or for new installations. Institutional investors’ attention tends to shift to other opportunities outside the industry, too. As a consequence, the rate of innovation begins to decline and once this happens consumers are to turn away from the industry’s products sooner or later since there are other, more attractive opportunities to spend money. Now, the industry begins, metaphorically speaking, to die - i.e. its share in the overall value added that an economy produces declines and its relevance gradually decreases.

Against this backdrop, one finds industry lifecycle theory in analogy to the lifecycle entropy of individual biological beings (Wuketits 1986: 180 ff.). The entropic moment enters the Gort-Klepper framework through the assumption that imitation of state-of-the-art products and processes becomes more difficult over time. It rests on the idea that the focus of relevant technological and commercial knowledge shifts from explicit to implicit or to “tacit” knowledge (Polanyi 1966).

However, Fehl maintained that within open social systems any entropic tendency in reality is countervailed by human creativity which, through innovative market forces and entrepreneurial arbitrage, injects new diversity and keeps the market process far away from any (entropic) equilibrium (Fehl 1983: 78 ff.). In this view, economic evolution is a principally open future process which follows no determined course (Fehl 2005: 82f.). A comprehensive theory of economic development that rests on the idea of economic development as a principally open future process which is driven by the pervasive activation of entrepreneurial and dis-equilibrating forces has been developed by Albert O. Hirschman (Pies 2006: 4). In the following paragraphs we propose to adapt Hirschman’s theory of economic development to the level of the evolution of specific industries - i.e. the world passenger car industry as case of reference - and to compare it to the Gort-Klepper type industry lifecycle.

2.2 Hirschman’s alternative approach to industrial development

Based on the approach of Gort and Klepper, the historic evolution of a number of industries has been analyzed so far, mostly in the form specific country studies. Among these are analyses of the US and the German automotive industry (e.g. Klepper / Simons 1997; Dreßler 2006). This article chooses an alternative, cross country perspective. By doing so, it opens the field for an application of Hirschman’s theory of economic development to a specific industry.

5

Albert O. Hirschman’s theory maintains that the development of industrial activities in certain countries or regions shows a different historic Gestalt depending on whether the country / region is subject to early, late or very late industrialization (Hirschman 1958: 8 ff.; Hirschman 1968). Early industrialization (e.g. in historic England) rely on the rather cumbersome process of accumulating an initial capital stock and attaining basic knowledge about how to run industrial processes. Late industrializing countries (e.g. historic Germany or the USA) dispense of a good part of the pre-work done in early industrializing countries and resort to experiences already made by others. Late developers oftentimes are able to jump with both feet in advanced technologies and manufacturing processes while early movers remain bogged down to activities that have passed their prime. Hence, late industrialization may unfold with particular élan and progressiveness. Finally, there are the very late industrializing countries (e.g. today’s emerging markets in Asia). Here, the process is dominated by import substitution, i.e. the substitution of imported manufactured goods by domestically produced goods.

It is important to understand that industrialization through import substitution (ISI) typically begins with consumer goods. Import substitution of consumer goods, however, in most cases rests on imported inputs and machines. The entire affair of ISI is, according to Hirschman, much more sequential and tightly staged than early or late industrialization. The ISI-process is comparatively smoother, less disruptive and less learning intensive than other ways to industrialize.

Hirschman had in mind the industrialization of entire economies. We try to adapt his theory to the purpose of analyzing the lifecycle of specific industries like the automotive sector. It is justified to make this transposition since Hirschman in fact explored the meso-level in economics, that is the dynamics of sub-sets of interacting agents (Calafati 2000: 7 f.). Mesoeconomics in this respect are mostly applied to analyze the distribution of incentives and development impulses between different groups of economic actors either within a geographical or a sectoral perspective. Hence, Hirschman’s paradigm of thought may be used to explore the historic evolution of specific industries.

The Hirschman paradigm of the industry lifecycle is sequenced as follows: Early entrants (e.g. the European pioneer companies in the automotive industry) invent the product and build up the initial manufacturing processes. Late entrants (e.g. American car manufacturers) build on this basis but are more bullish on advancing productivity and on expanding the market. Very late industry entrants (e.g. Asian car manufacturers) finally adopt the achievements of their predecessors through ISI and eventually apply them to new markets. The Hirschman paradigm contrasts the Gort-Klepper paradigm: The latter paradigm much more emphasizes the role of tacit knowledge (whose accumulation is biased in favor of early entrants). Since late and very late entrants tend to be less efficient than their predecessors, the know-how of producing and marketing any specific good increases incrementally over time but it also becomes more rigid and reinforces established ways of doing things. In other words, increasing efficiency and early success are being rewarded and reinforced under the Gort-Klepper paradigm but simultaneously lay the ground for technical and competitive conservatism at later stages of the lifecycle. Imitation becomes more expensive as time elapses and the pool of potential new entrants decreases successively. The Hirschman paradigm is much more open in this respect and emphasizes the role of imitation for industrial development. Imitation of products and processes established elsewhere keeps the industry lifecycle open for further development. In a free trade world, imitation by new entrants may also evoke new competitive pressure on incumbents even at historically late stages of the industry’s time path - either through price pressure or superior product policies. Good examples for this are the Japanese and Korean motor manufacturers who have seriously challenged the market position of European and American incumbents since the 1970s (Maynard 2003; Neumann 1996). This occurrence has not only widened

6

consumers’ choice but also spurred incumbents’ efforts and occasionally renewed their competitive élan (Ingrassia / White 1995).

Hence, as long as markets are open, there are countervailing, cross-regional powers to the maturing of an industry and there are forces which inject new levels of competitive diversity. This effectively reinforces the respective industry’s overall prosperity - irrespective of the stage of its lifecycle. There is no ex ante determined “rise and fall” under the Hirschman paradigm. Instead it rests on the assumption that imitation is easy and not too costly and that there is a pool of potential new market entrants.

7

3 Lifecycles in the world automotive industry

In the next paragraphs of this paper we will illustrate long-term developments in the world market for passenger cars. Main indicators are market entries and market exits and a rough evaluation of productivity trends in passenger car production. Finally we will present some thoughts on the ease of imitation and the reservoir of potential market entrants. We look at three market regions: Europe, The Americas and Asia (including Oceania). The analysis of market entries and market exits is based on the author’s database which compiles respective historic data on a company level. It focuses on manufacturers of passenger cars. Included are only those manufacturers who act independently in the market place, i.e. manufacturers which are not dominated by a concern or trust. Mergers are treated like a market exit of the company being taken over. A further criterion for the inclusion of a company in the database is a certain degree of technical independence, e.g. mere assemblers of CKD-kits or manufacturers of “replicas” and the like are not covered. We look at a historic period from 1888 until 2004/05.

3.1 Europe

The passenger car is basically a European invention and first efforts to bring it to market appeared here. The evolution of the number of European manufacturers is illustrated by figure 1.

Figure 1: Development of the number of passenger car manufacturers in Europe

A few motor manufacturers appeared in Europe in the second half of the 1880s but for some years their number reached not even a dozen . After 1895, however, the number of European passenger car manufacturers virtually exploded (“take-off”) before it consolidated at around 200 during the years preceding Word War I . There was another boost in the 1920s to more than 300 . Then, the number of European passenger car manufacturers steeply declined (“industrial shake-out”) to a range of 70 to 80 in the 1940s and 1950s and further though less steeply to approx. 40 to 50 in the 1970s, 1980s and 1990s and finally to 20 at the beginning of the 21st century .

The post World War I spike in the number of European manufacturers is an interesting caveat. It may be explained by the necessity of converting the war economy to peace time conditions. After World War I had ended, many industrial producers faced substantial overcapacities which they tried to fill with new products. Hence, some of them turned towards

8

manufacturing passenger cars - which had been a promising new business before 1914 (Edelmann 1989: 27). In Germany, the number of manufacturers was further propped up due to the hyper-inflation of the early 1920s which induced investment in real assets at a pathologic scale (Stucken 1964: 39; Stolper 1966: 70; Schultz 1976: 113; Blaich 1994: 24). Both, the conversion-boom and the inflation-boom of the European automotive industry finally turned out to be not sustainable. The post World War I spike was something pathological and ephemeral. It has not much to do with the general lifecycle of the passenger car industry in Europe. Leaving aside phase one may detect a pattern in European market entries and exits that reflects a full Gort-Klepper industry lifecycle.

Figure 2: Evolution of annual net market entry rates in Europe

Figure 2 illustrates the historic evolution of net market entries (net market entries of period t = market entries of period t minus market exits of period t) in Europe. Again, main characteristics of the Gort-Klepper paradigm become clearly visible: First the influx of new manufacturers and a positive entry rate at the beginning of the 20th century, later the beginning decline of net new entries before World War I. Apart from the ephemeral post-war new entry boom, the net market entry rate was negative throughout the next two decades. Thereafter it hovered around the zero line.

Gort and Klepper hypothesized that over time early market entrants are the more efficient producers. Hence, the probability to stay in business over the industry lifecycle should be higher for early entrants than for late entrants. Table 1 displays the average probability to terminate operations (“mortality”) in Europe. Rows Q indicate the year of market entry after the first market entry had appeared. Columns E indicate the years of operation after market entry. E.g. a manufacturer who entered the European passenger car business between year 0 and year 9 (Q0-9) after the first market entry had occurred faced a probability of 0.8% to drop out of operation within the following first 10 years (E0-9). Likewise, a manufacturer who entered the business in years 30 - 39 after the first market-entry had occurred faced a probability of 13.9% (Q30-39 / E0-9) to drop out during the first decade following his market entry.

9

Table 1: Average probability of market-exit by decade after market-entry (Europe)

Mortality of market entrants in cohort Q0-9 historically turned out to reach its first peak only after five decades (E40-49) at 5.5%. Mortality in cohort Q10-19 peaked at 7.9% after four decades (E30-39), in cohort Q20-29 after three decades at 11.4%. In cohort Q30-39 mortality peaked in the first decade (E0-9) at 13.9%. Thus, the later a manufacturer entered the passenger car business, the higher was the historic probability to fail early after market entry. Historic mortality in the European passenger car industry therefore appears to be very much in line with the pattern prediction of the Gort-Klepper paradigm. The later market entry took place, the earlier peaked mortality and the higher was peak-level mortality (with the only exception that mortality of new entrants in cohorts Q40-49 ff. tended to be lower again).

3.2 The Americas

Figure 3: Development of the number of passenger car manufacturers in America

Figure 3 illustrates how the number of American manufacturers evolved over time. The pretake-off phase began almost a decade later but was a little shorter than in Europe (~5 years compared to ~8 years).2 It was succeeded by a take-off phase which lasted from ~1898 until

² The American passenger car industry is more or less an exclusive North American story. Motor vehicle manufacturing developed into an important industry in Latin America after World War II, too, but for its major

10

~1913 and a phase of consolidation from ~1912-1924 . Then, the “shake-out” began which lasted until the 1940/50s. The number of manufacturers tended to decline further though much less steeply during the following decades .

The overall shapes of the curves displayed in figure 1 (Europe) and figure 3 are rather similar (except for the pathological post World War I peak in Europe). Also the sinus-like trend shapes of the net-entry curve indicate that the overall historic patterns of industry evolution in Europe and America are similar (cf. figure 4). What differs significantly are peak numbers of market entries which in America are substantially lower than in Europe.

Figure 4: Evolution of annual net market entry rates in America

Another difference becomes apparent when historic probabilities of market exit or “mortalities” are calculated for America (cf. table 2). New market entrants in America showed a tendency to be forced out of business earlier and with a higher probability than in Europe. Mortality of market entrants in cohort Q0-9 peaked at 11% after four decades (Europe: 5.5% after five decades). Mortality in cohort Q10-19 reached its highest level at 8.8% after three decades (Europe: 7.9% after four decades), in cohort Q20-29 at 13.4% after two decades (Europe: 11.4% after two decades). In cohort Q30-39 it climbed to 13.5 % in the first decade (Europe: 13.9% including the pathologic post World War I peak) and peaked in the second decade at 23.5% (Europe: 11.6%). Thus, the historic development of the American passenger car industry apparently was more disruptive than in Europe.

part, this was not born by indigenous manufacturers but rather by subsidiaries or licensees of US and European majors.

11

Table 2: Average probability of market-exit by decade after market-entry (America)

At first sight, the historic patterns of evolution which the passenger car industries in Europe and America show are very much of a Gort-Klepper type: The initial phase of market creation began in the late 19 ^th century (I) (II) Take-off appeared until 1910

(III) The number of manufacturers then stabilized until World War I (IV) “Shake-out” occurred between the wars and leveled off after World War II. (V) Consolidation set in during the 1960s.

What, on the other side, reminds the analyst strongly of the Hirschman paradigm is the fact that the number of market entrants in America was much lower than in Europe and that the disruptiveness of industry evolution was comparatively stronger. These results also reflect that the American passenger car industry was quite fast to develop large firms which reaped economies of size through mass production early. Passenger car manufacturing in Europe remained attached to handicraft or “custom and batch” manufacturing much longer, i.e. well into the 1920s (Scranton 1991; Braun 1997).

3.3 Asia

The historic pattern of the Asian passenger car industry (which is illustrated by figure 5) differs remarkably from Europe and the Americas. Europe was early to develop a passenger car industry and America comparatively late. Asia is a very late developer in this respect. The first few market entries appeared in Japan in 1912 (Nissan) and 1917 (Mitsubishi), some more during the 1930s (Toyota, Isuzu) but 40 to 50 years elapsed after the first entries until the number of automotive players grew significantly and fast.

12

Figure 5: Development of the number of passenger car manufacturers in Asia

In Asia, the pre-take-off-phase was extremely long and the passenger car industry’s actual take-off happened after post World War II.3 The 1970s and 1980s appear to be a first phase of consolidation with the net entry rate hovering around zero . Then, however, the number of Asian passenger car manufacturers further increased in the 1990s - mainly driven by Chinese new entrants . The number of Japanese and Korean manufacturers began to decrease towards the end of the 20th century due to several M&A deals (e.g. absorption of Nissan into Renault, Mazda into Ford, Mitsubishi into DaimlerChrysler, Daewoo into General Motors or Kia into Hyundai) but yet there has been no industrial melt-down or “shake-out” in Asia like the ones experienced in Europe or America between the wars. The peak-number number of market entrants in Asia so far has been much lower than in Europe or America. The sinus-like wave pattern of net entries is absent (cf. figure 6).

Figure 6: Evolution of annual net market entry rates in Asia

³ The number of manufacturers peaked in Japan in 1962 with the entry of Honda and in 1965 in South Korea. In India and China, passenger car manufacturing began as late as in the 1940s and 1950s. Iran began to develop its passenger car industry in the 1960s.

13

Table 3 displays the results of a mortality analysis for the Asian passenger car industry. Compared to Europe and America, historic mortality rates in Asia are low and they indicate a so far very much less disruptive evolution of the industry. Furthermore, there is no real evidence in Asia for the mortality of late entrants being higher than the mortality of early entrants.

Table3: Average probability of market-exit by decade after market-entry (Asia)

Relating these findings about the Asian passenger car industry to the historic developments in Europe and America, figure 6 demonstrates that since the 1950s Asian market entries numerically compensated for more than one third of the market exits in Europe and America and - on a worldwide scale - tended to countervail losses in manufacturer diversity during the last decades.

Figure 7: Development of the net gains / losses in the number of passenger car manufacturers

To challenge this view, one could stress the point that the late emergence and statistical longevity of Asian manufacturers has something to do with protection, state-ownership and strategic MITI-type trade policies. However, it is quite clear that protectionism and stateownership do not pay-off in the long run since they tend to keep productivity levels low (Alchian 1977: 137 ff.; Bhagwati 2007). Furthermore, there are serious doubts whether MITItype trade policies are the key to explaining industrial development in Asia (Bender 1997). It

14

is much more convincing to associate the above findings on the Asian passenger car industry with the Hirschman paradigm of lifecycle development. What matters most in this respect: Very late entrants may adopt existing products and process technologies easier than late entrants and early entrants (who hardly found any existing technology to adopt). The comparative ease of learning and imitation, which the Hirschman paradigm suggests, may smooth the disruptiveness of industrial development since, based on the experience made by others, it can limit the likelihood of mal-investment and bring new ventures quite fast to stateof-the-art levels of productivity.

15

4 Learning effects in the world passenger car industry

To measure learning effects in the world passenger car industry we use the admittedly rough indicator of average output per manufacturer. This is a first approximation since much more statistical research needs to be done on the early periods of the automotive industry’s history in order to attain appropriate indicators of finer granularity. Nevertheless, assume that the average output per manufacturer is a proxy for the representative firm’s productivity and its ability to use economies of size.

Figure 8: Logarithmic measurement of output per manufacturer (log. million passenger cars per manufacturer and year)

Figure 8 displays the logarithmic measurement of output per manufacturer in America, Asia and Europe. In the very early years of the 20th century, European manufacturers fell slightly but not seriously behind their American peers. ⁴ During the following approx. 20 years, however, the American gradient sloped much steeper than the European one. Output per manufacturer increased faster and a productivity gap opened between America and Europe (illustrated by figure 9). The gap widened until the 1920s and then began to close again over a time-span of approx. 60 years. Since around 1980, the gap’s size has returned again to a rather insignificant size.

For an explanation of this pattern recall Hirschman’s hypothesis that late entrants employ more advanced processes while early entrants oftentimes cling to the established ways of doing things. The development of output per manufacturer in America reflects a respective fast-track turn of major producers to advanced methods of mass production while the lagging performance of their European peers reflects their widespread contemporary affiliation to “custom and batch”. The gap began to close when general socio-economic developments in Europe favored a broader scope of individual motorization. A larger potential market size began to allow for more output per manufacturer.

⁴ Most probably, potential productivity was more or less equal in both regions but that the European industry was more fractioned country-wise.

16

Figure 9: Output per manufacturer gaps in the world passenger car industry (differences in log. million passenger cars per manufacturer and year)

Furthermore, one may detect feedback effects between productivity developments in America and Europe. European passenger car manufacturer strived to learn from their American peers when cheap imports from overseas put them under competitive pressure in their home markets. ⁵ Numerous “pilgrimages” of European automotive executives to Ford’s River Rouge plant are an eloquent reflection of their need to learn about advanced production processes. Output per Asian passenger manufacturer may be traced back into the 1940s when it was much lower than in Europe and in America. However, it increased exceptionally fast during the following years. Merely 20 years elapsed until Asian firms caught up with their European peers in the late 1960s. Apparently, Asia’s very late entrants learned comparatively fast about efficient ways of producing passenger cars and quickly adopted state-of-the-art technology. Figure 9 and figure 10 draw a more detailed picture which suggests that the process of “Hirschmanian” industry development is not over yet.

⁵ A good example may be found in German automotive history. In 1925, the German government began to gradually lift protectionist import barriers for passenger cars. As a consequence, the technically simple but affordable and comfortable six cylinder sedans manufactured at that time by General Motors and other American brands gained popularity among German business people during the “Roaring Twenties”. At the same time, sales volumes and profitability of incumbent brands came under commercial pressure since business people and self employed professionals constituted their most important customer segment (Edelmann 1989: 61ff.).

17

Figure 10: Logarithmic measurement of output per manufacturer in Western Europe and selected Asian countries (log. million passenger cars per manufacturer and year)

So far, Japanese manufacturers have been leading since the 1960s when domestic output per manufacturer caught up with European standards. Chinese and Indian manufacturers are still under way to catch up with their more advanced peers. In both countries, passenger car manufacturing started in the 1940s or 1950s. In India, output per manufacturer initially was comparable to the levels attained in Japan at that time. However, inward looking, autarky driven economic policies as well as thorough regulation of the economy retarded advances in productivity for many years after World War II. In the automotive industry, average output per manufacturer is still backlogged though data indicate that respective gaps are gradually closing.

Figure 11: Gaps in average output per manufacturer in selected Asian countries (differences in log. million passenger cars per manufacturer and year)

5 Summing up

The pivotal difference between the Gort-Klepper approach to industrial lifecycles and Hirschman’s theory of economic development respectively its application to the evolution of specific industries are the underlying assumptions about how easy it is to imitate state-of-theart production processes. Gort and Klepper maintain that imitation is easy at the early stages of an industry’s history and difficult at later ones. Thus, they are able to deduct an intriguing cyclic pattern of emergence, rise, maturing and decay which finds its analogy in the lifecycle entropy of individual biological beings. Indeed, large parts of the present article’s findings about the long-term evolution of the passenger car industry in Europe and America appear to be commensurate with the Gort-Klepper type industry lifecycle. It undoubtedly has its merits for explaining the origination of markets and industries as well as for explaining industrial “shake-outs”. Nevertheless, for comprehensively explaining the factual historic sequencing of the passenger car industry in a worldwide perspective, the application of Hirschman’s development theory to specific industries is more convincing. In a pure Gort-Klepper world one would not expect late developers to leap-frog like the American passenger car industry did in the early 20 ^th century. Equally, one would not have expected the ISI-style development which occurred with the emergence of the Asian passenger car industry.

19

References

Abernathy, William J.; The Productivity Dilemma - Roadblock to Innovation in the Automobile Industry (John Hopkins University Press, Baltimore and London, 1978).

Alchian, Armen A.; Economic Forces at Work (Liberty Press, Indianapolis 1977).

Bender, Dieter; Marktwirtschaftliche Reformprozesse und Wachstumsdynamik in Südostasien, in: Spiridon Paraskewopoulos (ed.), Wirtschaftsordnung und wirtschaftliche Entwicklung (Lucius & Lucius, Stuttgart 1997), pp. 301-322.

Bhagwati, Jagdish; In Defense of Globalization (Oxford University Press, New York 2007).

Blaich, Fritz; Der Schwarze Freitag - Inflation und Wirtschaftskrise (Deutscher Taschenbuch Verlag, München, 1994).

Braun, Hans Joachim; Konstruktion, Dekonstruktion und der Aufbau technischer Systeme zwischen 1914 und 1945, in: Hans Joachim Braun / Walter Kaiser (Hrsg.), Propyläen Technikgeschichte Band 5: Energiewirtschaft, Automatisierung, Information (Propyläen Verlag, Berlin 1997), pp. 11-282.

Burns, Arthur F.; Production Trends in the United States since 1870 (Augustus M. Kelley, New York, 1934).

Calafati, Antonio G.; Albert O. Hirschman on Economic Evolution (Università degli Studi di Ancona, Dipartimento di Economia, 2000).

Dreßler, Kristina; Der Lebenszyklus der deutschen Automobilindustrie (Josef Eul Verlag, Lohmar and Cologne, 2006).

Edelmann, Heidrun; Vom Luxusgut zum Gebrauchsgegenstand (Verband der Automobilindustrie, Frankfurt am Main, 1989).

Fabricant, Solomon; The Output of Manufacturing Industries 1899-1933 (National Bureau of Economic Research, New York, 1940).

Fehl, Ulrich; Die Theorie dissipativer Strukturen als Ansatzpunkt für die Analyse von Innovationsproblemen in alternativen Wirtschaftsordnungen, in: Schüller; Alfred, Leipold, Helmut & Hamel, Hannelore (eds.), Innovationsprobleme in Ost und West (Gustav Fischer Verlag, Stuttgart & New York, 1983).

Fehl, Ulrich; Wettbewerbliche Dimensionen des Oligopolmarktes, in: ORDO (Vol. 37, 1986a, pp. 141-153).

Fehl, Ulrich; Spontaneous Order and the Subjectivity of Expectations: A Contribution to the Lachmann-O’Driscoll Problem, in: Kirzner, Israel M. (ed.), Subjectivism, Intelligibility and Economic Understanding (Palgrave, Houndsmill et al., 1986b, pp. 72-86).

Fehl, Ulrich; Warum Evolutorische Ökonomik?, in: ORDO (Vol. 56, 2005, pp. 77-93).

Freeman, Chris; Soete, Luc; The Economics of Industrial Innovation, 3rd edition (The MIT Press, Cambridge, Mass., 1997)

20

Gort, Michael / Steven Klepper; Time paths in the diffusion of product innovations, in: The Economic Journal (Vol. 92, 1982, pp. 630-653).

Hirschman, Albert O.; The Strategy of Economic Development (Yale University Press, New Haven and London).

Hirschman, Albert O.; The Political Economy of Import-Substitution Industrialization in Latin America, in: The Quarterly Journal of Economics (Vol. LXXXII, 1968, p. 1-32).

Ingrassia, Paul / White, Joseph B.; Comeback - The Fall and Rise of the American Automobile Industry (Touchstone, New York 1995).

Klepper, Steven & Simons; Kenneth L.; Technological Extinctions of Industrial Firms: An Inquiry into their Nature and Causes, in: Industrial and Corporate Change (Vol. 6, 1997, p. 379-460).

Kuznets, Simon S., Secular Movements in Production and Prices (Kelley, New York, 1967).

Maynard, Micheline; The End of Detroit (Currency Doubleday, New York et al. 2003).

Möser, Kurt; Geschichte des Autos (Campus Verlag, Frankfurt / Main and New York 2002).

Nefiodow, Leo A.; „Der Sechste Kondratieff - Die großen neuen Märkte“, in Kondratieffs Zyklen der Wirtschaft, edited by Hans Thomas & Leo A. Nefiodow (Busse Seewald, Herford 1998), pp. 155-198.

Neumann, Horst; Mythos Japan (edition sigma, Berlin 1996).

Pies, Ingo (ed.), Albert Hirschmans grenzüberschreitende Ökonomik (Mohr Siebeck, Tübingen 2006).

Polanyi, Michael, The Tacit Dimension (Routledge und Kegan Paul, London 1996).

Schultz, Bruno; Kleine deutsche Geldgeschichte des 19. und 20. Jahrhunderts (Duncker & Humblot, Berlin, 1976).

Schumpeter, Josef A.; History of Economic Analysis, (Oxford University Press, Oxford, 1954), p. 1158.

Scranton, Philip (1991), Diversity in diversity: flexible production and American industrialization, 1880 - 1930, in: Business History Review (Vol. 65, pp. 27-90).

Stolper, Gustav; Deutsche Wirtschaft seit 1870 (J.C.B. Mohr - Paul Siebeck, Tübingen 1966).

Stucken, Rudolf; Deutsche Geld- und Kreditpolitik (J.C.B. Mohr - Paul Siebeck, Tübingen 1964).

Wuketitis, Franz; Schlüssel zur Biologie (Econ Verlag, Düsseldorf and Wien 1986).

21