History 1: The Rise of Oligopoly
The Early Years: Entrepreneurs and New Venture
The concept of automobility dates back centuries, figuring in the drawings of Roger Bacon in the 13th century and Leonardo da Vinci in the 16th. By the mid-19th century, various ventures produced vehicles that freed steam engines from their rails, albeit with little initial success outside of steam tractors and threshers. Meanwhile by the late 1800s streetcars transitioned from horse-drawn to electric, improving speed and lowering costs; the start of suburbanization predated motor vehicles. Finally came the carriage trade and the new-fangled development of bicycles. It was the German, Karl Benz, who in 1886 produced the first widely recognized car, which used a two-stroke internal combustion engine (ICE) that he had worked to perfect. Over the next 25 years perhaps 3,000 entrepreneurs tried their hand at making vehicles, particularly in France and then in the United States, but also in Germany, England, and Italy. Early vehicles took a wide variety of forms, with motorcycles and tricycles building upon the bicycle industry, and “horseless carriages” modifying four-wheeled horsedrawn buggies. Motive force came from steam, electric motors and lastly ICEs. Steam offered power but was slow to start and heavy. Gasoline engines were primitive, noisy, weak in power and prone to breaking down. Electric vehicles started immediately and were quiet, but had limited range.
Business concepts likewise varied. One market was for the wealthy to promenade their contraptions. Electric vehicles were suited to this and to other uses that revolved around a fixed base. In the United States, electric vehicles were the largest segment through 1903. Taxis and buses were another market. Into the 1910s most taxis in New York City were electric. The more technically inclined might engage in touring, or hire driver-mechanics for support. Racing was another early market, attracting large crowds and making the front page of newspapers. Again, electric and steam vehicles held speed records through 1902, when the wealthy American William Vanderbilt set a record in France. This was then bettered by Henry Ford in 1904 in a gasoline racer of his own construction. But overall reliability was poor, and production volumes miniscule, under 20,000 globally in 1901, with France the largest producer. Scientific American concluded in an 1897 article:
It is quite possible that the wonderful rapidity with which, in these days, a useful invention has developed from a crude idea into a practical shape with a positive commercial value has made us a little too exacting. . . . The recent competitions have proved that the perfected car, considered as a commercial product and something more than a mere toy, has probably yet to be built. . . . among the certainties of the future are a motor car which will shall be light, swift, strong, durable and cheap, which, as a means of banishing the noise and unsanitary filth of horse-propelled vehicles from our streets, and as a means of transportation for freight and passengers in country districts, will be as indispensable to the everyday life of the races as are the steam locomotive and the electric car [trolley] of our day.
“The Future of the Motor Car.” (July 17, 1897) Scientific American 77:3, 35.
Nevertheless there was widespread interest in this new set of technologies, an interest that rivaled that of electricity and railroads. Scanning the contents of Scientific American shows that in the early 1890s, motor vehicles were not yet far enough along to be more than curiosities. By the late 1890s, however, there was a regular column reporting novel vehicle configurations, speed and distance records and improved propulsion systems. At the turn of the century, cars were second to matters electrical, overtaking railroads, steel, and ships (Table 2.1).
Table 2.1. Number of Articles by Topic in Scientific American
All of this is reminiscent of other startup industries, as traced by Hannan and Carroll (1995) and subsequent studies in the “population ecology” tradition. From an early start in Germany, the core of the industry moved to France and then after 1905 to the United States. Rampant entry resulted in an industry peaking at about 300 firms in 1910 in the United States. Records are available to allow at least 770 firms to be tracked that actually produced a vehicle. The numbers in France, Germany, Italy, and the U.K. were smaller, and peaked closer to 1925. Over time there were over 300 entrants in Italy, France, and the UK, and 200 in Germany. Overall history documents 2,500 entrepreneurs who lasted long enough to sell a vehicle, but many more disappeared without a trace or never made it to market.
The venture capital phenomenon that we see today is nothing new. Henry Ford’s first two firms failed, leaving outside investors to foot the bills while he kept the engineering drawings. Then there is General Motors, which at the start was a stock market play engineered by Billy Durant, who had built a fortune through what we would now call financial engineering. The entity’s name reflected exactly what it was: an amalgam of firms that assembled cars or supplied parts to assemblers. Durant was, however, no manager, and the firm went through two early restructurings – one in 1909, when the bankers on the board exercised power to fire Durant after he was unable to come up with the financing to buy the Ford Motor Company, and the other in 1920 when Durant, back as president, again ran the company into the ground. There were likewise groups of serial entrepreneurs and managers, who moved from venture to venture. Durant himself was involved in a successful startup, Chevrolet, which he used to reacquire General Motors, and in Durant Motors, which failed. The remnants of Henry Ford’s second venture became Cadillac. Walter Chrysler first managed Buick, a GM subsidiary, then moved to Willys Overland (the forerunner of Jeep) and Maxwell before establishing Chrysler. Similar cases are found in Europe. In Germany, for example, Jørgen Rasmussen and August Horch were both involved in multiple ventures, including the forerunner of Audi. Ferdinand Porsche worked at Daimler and Audi and designed the VW Beetle. In Italy, members of the Ceirano family were involved with no less than nine different automotive ventures, including the predecessor to Fiat. In general, having executives with previous industry experience increased the longevity of these early ventures. The emergence of Detroit as the dominant production center in the United States was in large part because of the strength of its networks of entrepreneurs and managers.
Motor Vehicles Take Form
The emergence of successful firms took three decades, as entrepreneurs searched for appropriate products, learned how to make them, and found ways to get their vehicles from factory to purchaser.
Motive power was one early uncertainty. Many of the initial ventures sought to use steam, a rapidly maturing technology, as the motive force. For commercial vehicles tractors proved a viable market. Size was less of an issue, while reliability mattered. Case was the earliest of these, producing a steam-powered thresher in 1869, eventually becoming the world’s largest manufacturer of steam engines. For passenger vehicles, however, it took up to 30 minutes for a boiler to produce an adequate head of steam and they could explode in accidents. Nevertheless during 1898 through 1899 the Stanley brothers’ steamers were the best-selling vehicles in the United States, and for 5 years held the world record for the fastest mile.
By contrast electric vehicles were clean, quiet, and started instantly. The low-energy density of batteries of the time limited their range, while electricity was readily available only in cities. During 1900 to 1903, however, electric cars were the best sellers, and they remained in taxi fleets into the 1920s. Unlike steam and ICEs, in that era major improvements in batteries proved outside the realm of the feasible. Lithium batteries, for example, were only invented in the 1970s. While Porsche developed a gas-electric hybrid in 1899, that too proved a dead end. By 1905, the market for electric cars was clearly disappearing.
Successful ICEs dated to the work of Karl Benz, whose two-stroke engine initially used natural gas. Meanwhile Gottlieb Daimler and Wilhelm Maybach worked for Otto Langen, who developed an early four-stroke engine. They both left in 1883 to develop engines on their own. Initially their aim was the market for stationary engines, as ICEs were far smaller than steam ones, but also less powerful. They, however, also developed cars and licensed their engine to other entrepreneurs such as Panhard et Levassor in France. Noise, smoke, unreliability, and low power meant that until the early 1900s ICEs were used only in a minority of the cars being produced.
Progress in ICEs outpaced that of reducing the size of steam engines. In short order Benz, Robert Bosch, and others developed spark plugs, carburetors to create a good air–fuel mix, water cooling, and other advances that led to more compact, reliable, and powerful engines. Bosch’s high-voltage magneto and spark plug combination was particularly important, as it allowed a strong enough spark to ignite a fuel–air mixture at high speeds, boosting the power that could be produced by an ICE. Gasoline proved to have a high-energy density, unlike batteries. Combined with transmissions that could be shifted with a lever and clutch pedal—necessary because engines had little torque at low speeds—this made ICEs a viable alternative to steam. These various innovations were all commercialized by 1902, and by 1905 ICEs became the dominant power source for automobiles.
Meanwhile other developments focused on the physical structure of cars. Some early entrepreneurs came out of the carriage trade, or purchased and modified four-wheel buggies, which were readily available. These were tall, and passengers and the conductor or chauffeur sat on top of the engine. Since vehicles frequently broke down, the wealthy purchasers of early vehicles often employed others to keep them running. Other entrepreneurs came out of the bicycle industry; they tended to start with three-wheeled vehicles, such as Fiat, with the engine resting over an axle that connected two rear wheels. The modern car literally took shape with the French Panhard system that placed the engine at the front. Occupants were separated from the engine and sat lower to the ground, improving the ride. This architecture spread quickly, and by 1905 most cars no longer looked like horseless carriages. In Germany, though, three-wheeled vehicles remained the largest sellers by volume into the 1930s, and in Japan into the 1960s.
Purchasers: Who and How
The business model was a second area of innovation. In Europe—particularly France—passenger cars remained an elite product. Indeed, most 19th century cars were unique, rather than essentially identical versions of a base model. Early firms bought engines. For instance, Panhard used a Daimler engine under license, adding them to the chassis, and then had a coach builder add body and seats to their customer’s specification. This impeded production efficiency, but with a market aimed at high-income families, reputation and product differentiation increased sales, which because of high margins was more important strategically than lowering the manufacturing cost. In addition, in much of Europe alternative means of transport were relatively well developed. By contrast, in the United States, roads were often little more than wagon trails, unimproved except for the clearing of trees and rocks and the presence of bridges where fording a stream was impractical. While a venture’s initial vehicle might have been bare-boned, firms everywhere tended to move up-market, where margins were better. The U.S. economy was growing more quickly, and incomes were higher; there was a larger middle class and relative to Europe, many of these families lived outside cities large enough to have streetcars.
At the turn of the century, volumes remained tiny from today’s perspective. Oldsmobile and then Ford focused on more utilitarian vehicles, and on a single model. Nevertheless that meant that in 1902 Oldsmobile made only 2,500 of its Curved Dash model, the best-selling car in the world. The political fragmentation of Europe kept domestic markets small. While in the 1890s most cars in the United States were imported, high tariffs impeded exports while encouraging domestic production. The size, wealth, and structure of the American market meant that by 1910 assemblers in the United States attained production volumes far larger than those in Europe. The invention of the electric starter in 1911 further expanded the size of the market. It not only made vehicles safer, but without the need to crank an engine, women could more readily operate cars.
One component of the automotive business model was the structure of the value chain. What aspects of the business would car companies themselves undertake? In the early years, an entrepreneur would work with a vehicle concept, and purchase parts from a variety of sources, including firms that made engines—for example while Peugeot assembled motor vehicles, they also sold engines to other car companies. In the early years, this enabled entry, as an entrepreneur could buy everything from suppliers so needed little capital. Those who succeeded, however, also tended to have previous experience in the industry. Although volumes rose, at most U.S. firms production remained craft in nature until the 1920s, and in Europe until the 1930s. Those who worked at the factories of Oldsmobile or Citroën were skilled workers, expected to have their own tools which they used to fit parts. With a multiplicity of firms, in some cases with more than one model, parts firms too remained relatively unspecialized, as no single part could be made in significant volume. When the Society of Automotive Engineers was founded in New York City in 1905, the directors hoped that it could create standards for parts around which car companies could design their products. That would then lead to larger volumes and greater production consistency.
The growth of demand, particularly in the United States, allowed firms to gradually increase their overall sales and revolutionize production.The success of Henry Ford’s utilitarian Model T, launched in 1908, soon meant that demand was outpacing the rate at which he could assemble vehicles. One approach was to improve operational efficiency by introducing the assembly line, gradually implemented in 1913. This improved work flows and helped boost output, but without interchangeable parts assembly continued to depend upon experienced workers. But the skilled fitters of that day remained mobile. To obtain greater stability, Ford implemented the $5-day in 1914, and in 1926—helped by falling demand—also cut the standard work-week from 6 to 5 days. More important, Ford’s production staff under William Knudsen worked to standardize production, investing heavily in dedicated machine tools that could turn out parts more consistently. With closer tolerances, parts would no longer require skilled fitters. As that effort gained momentum, the assembly process was speeded up, and skilled workers shifted to maintaining the machinery that undergirded the new production system and similar tasks.
This new system also required discipline in the production process. In the past, parts were reworked until they fit. For the assembly line to flow smoothly, workers and their immediate managers had to have the authority to scrap parts that did not quite fit, rather than rework them. This went against all previous manufacturing logic, but as Ford Motor implemented these various pieces of the new mass production system, the company was able to simultaneously achieve production volumes unheard of for such a complicated product, lower costs through economies of scale and improve product quality. Ford was also forced to integrate vertically. To manufacture consistent parts and to keep up with the pace of growth required that “upstream” operations use specialized machine tools. As sales increased, Ford Motor also had the money to finance the associated investment. In Europe, Fiat was an early adopter of Ford’s approach, doing so in the 1910 to 1920 era, though with lower degrees of vertical integration. In France, Peugeot was alone in having a limited product line, but was unable to move away from craft assembly until after World War II (WWII). On the whole, the slower pace of growth and the greater variety of product meant that parts production in Europe tended to rely upon standardized machine tools with jigs and tooling that could be seamlessly switched. The same was true as the industry developed in Japan in the 1950s and early 1960s. In addition, motor vehicle manufacturers were less profitable—in Japan Toyota had to be restructured in 1950, while Nissan went through a costly strike in 1954. Assemblers simply did not have the financial resources to integrate vertically. But all understood the logic of longer production runs that would facilitate designing and manufacturing interchangeable parts, which in turn would enable more efficient assembly.
How to move the metal remained a challenge. Assemblers quickly hit upon the financial model: pay suppliers slowly, collect from customers quickly. As long as they remained primarily assemblers, motor vehicle manufacturers then incurred the costs of engineering and design, the expense of final assembly, and the budget for national marketing. If they could collect deposits when cars were ordered, and the balance upon delivery, they could pay suppliers out of the proceeds and needed relatively low capitalization. This was one factor behind the pace of entry, as an entrepreneur only needed a few well-off backers, and success meant the potential to offer very high dividends on the initial investment. Of course that worked best if a company could keep expanding, as the money flowing in would readily cover operational costs. A fall in sales would reverse that process, and quickly put a firm out of business—as happened to most entrants into the industry.
When sales were low, and the product aimed at the wealthy, sales agents could be used, while repairs were up to the owners. As the sales volume increased, customers were no longer uniformly wealthy, and assemblers turned to specialized dealerships that handled sales, delivery, might provide repairs and could even arrange finance. Even more important, as the market moved away from made-to-order vehicles, those selling cars needed to hold inventory. In the process, the role of dealers shifted: from the car company’s perspective it was the dealer, not the final customer, who was buying the vehicle.
Car companies did try to run their own stores, but these ran into two problems. First, dealerships were complicated businesses. Their revenue stream was a mix of margins on new vehicles, and income from service and repairs. By the 1920s, as more and more sales were to existing car owners, dealers also took trade-ins and sold used cars. Finally, they needed real estate, inventory, and local marketing. In most countries, these needs varied from town to town, while sales varied both over time and on a regional basis. Dealers needed considerable flexibility, whereas to keep operations tractible, car companies sought to impose standard operating systems.
Car companies have proved inept at juggling the multiple business lines that a dealership operates, from new and used cars to finance, service, and parts sales. Ford tried direct “factory” sales in the early years, as did other manufacturers in Europe and Japan. Toyota went through a restructuring in 1950 when its in-house distribution arm built up excess inventory going into a recession. Ford lost billions on its 1998 to 2001 “Auto Collection” experiment of factory stores in Tulsa, OK and other select markets. A manufacturer—the “factory”—is interested in steady sales volume; in-house managers with no equity in the business prove prone to holding excess inventory and offering discounts. Stand-alone franchises benefit from greater freedom to set compensation to performance metrics and to experiment with their business, and flexibly respond to local rivals and local sales conditions. So by 1930 the franchised dealer system was in place in the United States, and by 1960 in Europe and Japan. There is one common exception: “showcase” urban dealers are unusually expensive to operate but are viewed as crucial for marketing purposes. So either a franchised dealer needs to be subsidized, or they need to be run as money-losing “company stores.”
One final aspect was finance: retailing consumed huge amounts of capital, for real estate and especially for inventory. To give a sense of the order of magnitude, in 2016 dealers in the United States alone collectively held $208 billion in vehicle inventory, and billions of dollars of additional real estate. In contrast, the manufacturing end of the industry held only $35 billion in working inventory, and “just-in-time” manufacturing meant that much of this was held by parts suppliers. Quite simply, aside from boutique manufacturers such as Tesla, car companies cannot afford to sell their own vehicles directly to consumers.1 To do so would require the investment of more money in dealers than the total market value of any of today’s leading producers.
As the shape of the industry coalesced around vehicle designs and business models that remain dominant a century later, so did the wider social and legal network in which the industry is embedded take form. Proper usage was specified, with speed limits, lanes, and turning rules, while pedestrians were pushed from the street to sidewalks. Drivers were licensed and cars registered and taxed. Anti-theft mechanisms such as vehicle identification numbers (VINs) were institutionalized, insurance was made available, and a legal framework developed for dealing with the large number of motor vehicle accidents. Infrastructure was standardized, from traffic lights to gasoline stations. Professional associations developed road designs—curves, road slopes, limited intersections—that facilitated faster travel, while improving construction methods. Fuel taxes helped fund these changes, and particularly outside the United States high taxes shaped the nature of demand, though fuel efficiency did not become a regulatory target until the 1970s. Safety issues were present, with headlights, turn signals and by the 1920s, safety glass. Cars began to show up in movies and music. Local boosterism favored motor vehicles over streetcars, shifting mobility options. That in turn shifted decisions over where people would live and businesses would locate, in turn affecting patterns of usage. Over time these and other elements co-evolved, with private car ownership and operation at the household level at the center. This made it increasingly difficult to make other than incremental change to any single component.
Consolidation and Oligopoly
Technology continued to evolve during the era following World War I (WWI), but by 1939 virtually every element of a modern car was present, or had been used, though air conditioning, automatic transmissions, and car radios were not widespread until the 1950s. In the middle- and lower-priced end of the market, the cost advantages of mass production led to fewer producers. As cars became items of daily use, open carriages gave way to closed bodies. However, the crafting of these from a combination of wood and metal, and the slow process of painting, or rather of paint drying, continued to limit economies of scale. Furthermore, the long production runs that allowed Ford to lower cost had a downside. As the economy evolved, tastes and expectations changed as well: models with new styling and improved features eroded the sales of the Model T, paving the way for new entrants such as Chevrolet and the firms that evolved into Chrysler. But Ford’s biggest competitor was himself. The basic structure of the Model T was designed to be rugged and durable; its simplicity lent itself to ease of repair, and parts were available everywhere. The main differentiator between a new and a used car was thus price, and no matter how much Henry Ford lowered production costs, he was unable to sell a new car for less than the price of a used Model T. Lower production costs alone were insufficient to succeed in the industry, though the low costs that came with scale mattered.
Ford briefly held a near-monopoly on the U.S. market. Interchangeable parts and the production line allowed the company to lower the price of the Model T year after year, and sales exploded to peak at 80 percent of the market. As detailed by Mira Wilkins and Frank Hill in a seminal work on multinational business, American Business Abroad: Ford on Six Continents (1964), the company set up branch plants such as in Dagenham, England an integrated production facility that meant Ford briefly accounted for 60 percent of global sales. By 1910 new entry had already slowed, and in the 1910s firms exited the U.S. industry in increasing numbers. The 200-odd producers in 1910 fell to 150 by the end of WWI and to less than 100 by 1925.
As is often the case, monopoly paves the way for its own demise. Ford’s success rested on a strategy of the production of a single model that with hindsight was unsustainable. Success also bred a divorce from market realities. The company was closely held, and Henry Ford did not need to answer to anyone else. Over time he drove out most of the managers who had built his production system, ran his marketing, and interfaced with his dealers. As briefly the richest man in the world, he had no need of banks, and in an era before the imposition of corporate income taxes, saw no need for accountants. The company kept no books. As the very success of the Model T undermined sales, Henry Ford’s monomaniacal strategy and management opened the way for rivals. By 1927 sales collapsed, and the company was forced to cease production until it could come out with a new model. But Henry Ford remained at the head, and until the 1950s the company continued to struggle in third place, behind General Motors and Chrysler. Ford’s monopoly gave way to a three-firm oligopoly.
Advances in the steel industry resulted in a more important shift. In 1922 the Budd Company, an independent body-builder, used the availability of better sheet steel, deep-drawing stamping presses and improved welding methods to develop an all-steel enclosed body. This had three advantages. First, it facilitated more dramatic styling with more frequent changes. Second, it removed the need to deal with the vagaries of wood as a material, and provided an intrinsically quieter body. Finally, the entire body could be put in an oven, so painting a vehicle took hours, not days. This shifted the economies of scale of the entire production process, because the costs of stamping presses and paint ovens raised the capital expenditure needed to produce a vehicle, while lowering direct production costs. Dies, specific to a given model, were also expensive. A firm that could sell enough of a model enjoyed significantly lower production costs while producing a better-appearing vehicle. Few of the incumbent firms could afford the investment, and by 1929 the “Big Three” of General Motors, Chrysler, and Ford accounted for the overwhelming majority of industry sales. Although the timing differed—around 1950 in the various national markets in Europe and 1965 in Japan— the industry everywhere coalesced into an oligopoly dominated by three or so producers.
General Motors was particularly successful in this transition. One aspect was the development of modern branding, with an array of marques ranging for example from the entry-level Chevrolet through Pontiac, Buick, and Oldsmobile to Cadillac at the premium end. This product differentiation allowed GM to segment users and enhance profits. Second, GM learned how to share parts and even body architectures across brands and models, enhancing volumes and further increasing profits. Third, as detailed by Alfred Chandler, the founder of modern analytic business history, GM revolutionized the management of large, complex firms. The chemical giant DuPont had recently divided their operations into divisions organized around product lines. They were GM’s largest shareholder, and under Alfred Sloan, whom the DuPont family named as GM’s president. This approach was combined with more sophisticated accounting such as return-on-investment (ROI) calculations to balance divisional autonomy with accountability. Finally, a distinction between staff and line functions created a cohesive whole out of the disparate car assemblers and parts manufacturers that Billy Durant had cobbled together. Previously, parts operations did not necessarily supply the rest of GM, while the biggest competitor of one GM brand might be another within the company’s fold. The new structure addressed these inefficiencies, allowing General Motors to enjoy both the highest margins and the lowest production costs in the industry.
The incestuous nature of management of the Detroit Three facilitated their oligopoly. Although there might be limited mobility between engineers and managers at Ford and General Motors, they sat together on the boards of community organizations and were members of the Detroit Athletic Club. Workers at almost all their plants were members of the United Auto Workers, which as traced in Chapter 10 led to similar labor costs. While productivity continued to improve, technology otherwise evolved slowly and so innovation did not allow one firm to distinguish itself from another in costs or in product features.
The oligopoly was effective. During the 1960s General Motors was year after year the most profitable firm in the world, with a return on equity of 40 percent and a return on capital of 20 percent. These profits were shared under GM’s price leadership, backed by their well-founded fear that achieving more than a 50 percent market share would invite antitrust action by the US Department of Justice. Coordination was simple. Annual model changes the norm by the 1930s. With all of the Detroit Three seeking to build a similar set of low-to-high brands, GM could engage in price leadership. They would announce provisional prices first, and Ford and Chrysler would come in with slightly lower prices, gauged to let them capture half the market. GM would adjust final prices down if it judged the gap too wide. Only once, in 1955, did this break down into a price war. GM matched steep price cuts and kept market share. Ford and Chrysler gained nothing but lower profits. They learned their lesson, and thereafter none of the Detroit Three rocked the boat.
1 As of 2015, Tesla had wait lists for its vehicles. The firm’s capital needs will mushroom if sales slow, or if they move down market where carrying inventory will be more important.