In 1919 GM had been selling at one hundred dollars a share. By November 10, 1920, it hit fourteen dollars.37 In a brutal all-night session, Durant’s key investors—bankers from J. P. Morgan and Pierre S. du Pont, scion of the famous munitions firm—negotiated a buyout of Durant, who agreed to quit as president of GM. Du Pont agreed to take his place, and brought Sloan in as his assistant.
The situation was dire. Du Pont and Sloan had to turn the company around in a down market. They did have an inventory they could write off as a tax loss, and enough remaining cash to stay solvent for at least a year. Still, the livelihood of 85,000 workers was at stake—not to mention the portfolios of investors large and small. Alfred Sloan’s entire personal fortune was wrapped up in GM stock.38
The solution they came up with revolutionized American business. At its heart was a paradox. The best way to make General Motors a single integrated company, they decided, was to give each separate division, from Buick and Chevrolet to Pontiac and Oakland Trucks, as much freedom as possible. Let each division’s chief executive produce the products he saw as best suited to his share of the market, and best suited to his factories and engineers and workers. Then keep overall control strictly limited to coordinating the different divisions and making sure everyone was staying profitable.
In effect, Thomas Jefferson had been as right about business as he had been about political constitutions. The best corporate government was the one that governed least. “Decentralization [is] analogous to free enterprise,” Sloan later wrote, “Centralization, to regimentation”39 The best way to run a complex corporation was to have the boss at the top providing only an overall direction and oversight, while turning the entrepreneurial instincts of his top executives loose on the problem of how to produce cars and bring them to market. Meanwhile, the chief executive’s job was not to give orders but gather information, in order evaluate the company’s overall progress, anticipate problems, smooth out bottlenecks, and, as Bill Knudsen might have said, keep all noses pointed in the right direction.40
On New Year’s Day, 1922, Sloan presented his plan to the board of GM, which enthusiastically endorsed it.41 There was only one question. Would it work?
That depended on getting the right man to put the new GM corporate culture to the test. Sloan decided that was Big Bill Knudsen.
Almost as soon as he had finished sorting out the final details of the General Motors reorganization, he picked up the phone. A few days later, Knudsen was in his office. Sloan explained that he had no particular job at GM in mind for Knudsen, but that he needed someone good on his staff—“someone who can help the operating units do a better job. If you would like that job, we would like to have you here.”
“All right,” Knudsen said.
“How much shall we pay you, Mr. Knudsen?”
“Anything you like,” Knudsen replied with a shrug. “I’m not here to set a figure.”
“What were you getting at Ford?”
“Fifty thousand dollars a year,” came the answer.42
So Sloan started him on February 23, 1922, at six thousand dollars a year.
Knudsen didn’t care. It wasn’t where he started, but where he’d finish, that mattered.
As Sloan soon discovered, Knudsen could make almost anything out of metal. Since leaving Ford, he had already developed his own design of muffler that could be assembled from its simple parts in less than two minutes. He had taken it to the parts department at Chevrolet, where the purchasing agent, Donald O’Keefe, asked him what made it different. Knudsen picked it up and said, “It’s strong!” Then he smashed it bodily against the shop floor with a resounding clang. Chevrolet still turned him down.43 But a more enduring link between the Dane and the company was about to be forged.
Chevrolet was Sloan’s problem child. It was GM’s lowest-priced, but also least profitable, division. In 1921 Chevrolet had lost more than $8 million—the equivalent of almost $140 million today. Many doubted it would ever turn a profit again. Expert consultants Sloan had brought in argued that the line should simply be shut down. Sloan argued forcefully to Du Pont that this would be a mistake. With better engineering and a new head, Sloan believed Chevrolet could ultimately replace Ford as America’s choice for a low-cost, high-quality car. “Forget the reports,” Du Pont finally said. “See what we can do.”44 So in March 1922, Sloan turned the troubled company over to Knudsen.
A couple of weeks later, the same Donald O’Keefe was passing through the GM building hall when he spied the familiar figure of Knudsen. “Say,” O’Keefe said, “are you still trying to find someone at Chevrolet to buy your muffler?”
Knudsen smiled. “No,” he said in a husky half whisper, “I’m now the company vice president.”45
For some, the twenties were the era of Jazz Age parties, flappers, gangsters, and bathtub gin. For Sloan and Knudsen, it was an era of nine- to ten-hour workdays six days a week, of pouring over piles of balance sheets and inventory lists, long phone calls with suppliers and distributors and short memos to subordinates, as well as painstaking inspections of assembly lines.
Sloan knew if Knudsen could turn Chevrolet around, it would prove his new business philosophy was right. He and Du Pont were now paying the big Dane $50,000 a year on that bet.46 Knudsen, meanwhile, was determined to show that the continuous assembly line was not a Ford fluke but a way to turn any manufacturing company into a profit winner. “If we work hard,” he told Du Pont, “then we’ll get some business.”47 He became famous for wearing his hat in his Flint office because he never sat down long enough to take it off. He was always on his feet, throwing on a coat, and hurrying down to the various plants along Chevrolet Avenue to see what was happening—and to make sure his instructions were followed to the last screw, bolt, and casting.
Everyone who dealt with him noted Knudsen was a driven man, but a soft-spoken one. It went back to his days at Keim, where he had earned a reputation for a ferocious temper that, combined with his commanding figure and his skills as a boxer, made him a dangerous man to cross. One day an angry worker tried to do just that. Knudsen came back to the office after leaving the man flat on his face on the shop floor.
His mentor Bill Smith shook his head. “If you’re going to fight one,” he finally said, “are you going to be ready to fight them all?” Knudsen didn’t answer. Then Smith said something that Knudsen never forgot. He said, “From now on you’ve got to lead, not drive.”48
It was the last lesson Smith left him, and one of the most important. From that day Knudsen learned to bottle up his rage and lock it away, and deal with his employees, even the most menial, patiently as equals. “I learned when you shout at someone,” he once said, “you make him afraid. And when he’s afraid, he won’t tell you his troubles”—or tell a manager the truth about what on the assembly line wasn’t working, or what had gone wrong. An old Ford employee later put it this way. “Mr. Sorensen,” he said, meaning Knudsen’s rival Cast-Iron Charlie, “was a wild man, Mr. Knudsen a mild man.”49
Many years later, in wartime Washington, some people would mistake Knudsen’s gentle manner for meekness or a lack of conviction. It wasn’t, as employees and colleagues at Chevrolet had learned. It was the manner of a man who had learned that the price of being a true leader was self-mastery.
Knudsen kicked things off by breaking the production of Chevrolet into three stages. First came the castings, forgings, and stampings of steel and other materials, he explained to O’Keefe and his other managers. Then came the machining and assembly divided between five principal plants, with engines and major axle subassemblies being completed at Plants No. 2 and 4 and the chassis after 1923 being formed at Fisher Body’s new plant across the street. Finally came the finishing, including painting and undercoating.50
To deal with the first stage, Knudsen put all the ordering of Chevy’s primary materials—steel, iron, zinc, rubber, copper, and the rest—into O’Keefe’s hands, with all orders to be placed six to nine months ahead. At the same time, he lef
t the actual purchasing of those materials to the five principal plant managers, on the theory that they would know best where to buy, and find the best price. To deal with the second stage, Knudsen worked closely with machine tool manufacturers, converting the Chevrolet production line into using as many multipurpose tools as possible, in order to speed up tool ordering but also use floor space more efficiently.51
As for the final stage, when Knudsen had first come back to Buffalo to make Model T’s, he had taken away every file and hammer in the factory. He now did something similar in the Chevy plants, in order to train his workers to depend upon the machining process to make every part fit. The goal of mass production, as he never tired of explaining, was not to make things faster or to make them all look alike. It was to make them all work alike, so that every Chevrolet engine or transmission performed exactly like the previous one. That in turn required an accuracy measured down to the hundredth of an inch, which must elude even the most experienced craftsmen—and which only machines could provide.
“Speed produces nothing in manufacturing,” Knudsen liked to say—which was one reason he eschewed the complicated time-and-motion studies of production gurus like William Taylor. “Accuracy is the only straight line to great production.”52 Knudsen was moving automobile production out of the workshop mentality and into modern manufacturing.
With Knudsen in charge, Chevrolet made a sharp U-turn. Sales jumped from 72,000 to almost a quarter million cars and trucks in 1922, and the company that had lost $8.6 million when Knudsen stepped in showed a net income of $11.2 million.53 Pierre S. du Pont smiled, announced to Sloan that he would now be president of GM, and retired to his enormous home in Longwood Gardens to tend his orchids and figs.
The next year, Chevy sales and income jumped again. But there was more to come. In January 1924 two thousand of the company’s dealers met at the Palmer House in Chicago. Knudsen was there, and as he walked through the lobby into the ballroom, dealers came up to shake his hand and slap the back of the man who had saved their livelihoods and given their cars a future.
Knudsen made to sit at the head table, but his sales manager, Dick Grant, caught him by the arm.
“Mr. Knudsen,” he said, “you just can’t sit there and expect me to do all the talking. You’re the boss, so you have to say something to these people.”
Knudsen had never given a public speech in English in his life. He struggled to his feet before the cheering throng, his mind a complete blank. In desperation he glanced back at Grant. “What do I say to them?” he whispered.
“Whatever you’ve got in your head,” Grant hissed back encouragingly, his hand cupped against the side of his mouth.
Knudsen stared out across the sea of humanity, studded with waiters clearing dishes and the steady updrift of cigar smoke.
Then suddenly he raised both hands over his head, his index fingers extended.
“I vant vun for vun!!” he shouted in his thick Danish accent, and then abruptly sat down.54
At first there was a stunned silence. Then those who understood the message whispered it to others, and soon the entire room knew what Knudsen had meant. He wanted one Chevrolet to sell for every Ford Model T. It was the first open challenge by a car company to Ford’s sales supremacy, and the dealers loved it. Pandemonium broke loose. They cheered, they stamped their feet, they pounded the tables. Some stood on their chairs as they roared their approval. “Vun for vun” spread through the company and became the new battle cry for Chevrolet. Dealers, managers, and even workers were sensing that Knudsen was about to push them toward another major breakthrough.
And they were right.
The experts call it “flexible mass production”: a manufacturing process that allows for constant modification and change. This was the second revolution William Knudsen introduced to the auto industry, after the continuous assembly line he developed at Ford. Flexible mass production was embodied in the idea of the annual model. Billy Durant’s consumer-driven car culture was about to become reality.55
At first Sloan resisted it. The notion of having to redesign a perfectly fine car every year seemed wildly extravagant, and expensive. “We are all against yearly models,” Sloan confessed at the end of July 1925. But he added, “I don’t see just what can be done about it.”56
This was because Knudsen had already started the process with that year’s coming model. Even more important, by bringing in the dealers to tell him what customers wanted and how Chevrolet could make a better car, Knudsen had thrust the consumer deep into the production equation. The presumption was that by gratifying consumer preferences, as well as constantly improving engineering performance, GM could sell a new Chevy even to someone who already owned one—and who in turn would come to expect better and bigger changes the next year. The consumer culture’s “cult of the newer” was about to spread into the industrial economy—and Sloan wasn’t about to stand in the way if the idea also pulled the rest of his General Motors into the black.
Sloan gave the nod. The new Chevrolet would be launched in September 1925, to be available the following year. It’s the pattern all auto companies have followed, from that day to this. The 1926 Chevrolet was so much a hit that it slashed the Chevy-to-Model-T sales ratio from thirteen to one to two to one. The goal of “vun for vun” didn’t seem so far-fetched after all.57
Yet all this would have been moot if Knudsen didn’t know how to deliver the goods, by a rapid systematic retooling of the assembly line. “All old machines were discarded,” as Knudsen later described, “new heavy type standard machines (not single purpose) were installed, and fixtures strengthened as to withstand the spring, which is the greater factor [in causing inaccuracies] than wear.”58 Now Knudsen could set lower limits on the precision done by his grinders and borers and other tools, which meant less waste and more savings in raw materials.
It also meant his workers and managers could meet changes faster, because their tools were ready to adapt to new engineering demands. Knudsen focused Chevy’s satellite plants in Toledo, Detroit, and Bay City, Michigan, on specialty subassemblies and pushed final assembly out to places like St. Louis and Oakland, California, with a branch of Fisher Body at each. In effect, he had turned Sloan’s decentralization principle into a way to mobilize the entire division more efficiently.59
Knudsen’s next car, the 1927 Chevrolet, was an even bigger hit. More than a million cars were sold, and the impact was felt all the way to Dearborn. It was the ’27 Chevy that finally forced Henry Ford to abandon the Model T for a new car, the Model A.60 Eight years earlier Knudsen had told the old man he could make the changeover in under six months. Without Knudsen, Ford was compelled to shut down production for almost a year—a year that gave Knudsen room to spring his next big surprise.
It came when General Motors announced that the new 1929 Chevrolet would sport the first six-cylinder engine for a low-priced car, instead of the usual four. Knudsen and his engineers had been planning out the change almost two years earlier. They had even lengthened the wheel base of the 1928 model by six inches in order to accommodate the future bigger motor. A pilot plant was set up in Saginaw, while Flint continued to make the four-cylinder job. Over September and October, they made almost two hundred of the new engines, before moving every machine tool, jig, and fixture out to Flint—and ordering enough extra to produce six thousand engines a day.61
By December Flint was making two thousand a day, and Knudsen could sit back at last, knowing they would be up to speed with the new year. The result was that a couple of weeks after the inauguration of President Herbert Hoover, as customers across the country crowded into showrooms to look at the revolutionary six-cylinder Chevrolet, not one had to wait for the model or the color they wanted.62
Few new cars ever quite produced the sensation that the ’29 Chevy did, then or afterward. It came in no less than eight different models (from the two-door coupe to the four-door Cabriolet with a rumble seat in the back) and could drive at seventy miles per hour
all day—the first low-priced American car to do so. Best of all, thanks to Knudsen’s careful economies, it cost no more than the four-cylinder ’28 model: $495 for the coupe, and $695 for the standard four-door version.
Knudsen’s 1929 Chevrolet was the direct ancestor of the modern automobile. It was also the last to be introduced before the Depression. Its success enabled GM to weather the next several years of economic storms, as the slogan “Vun for vun” became a reality. In 1929 Knudsen could look at sales of an unprecedented 1.3 million cars and trucks; only Ford with his new Model A did better. Ford still held a lead in 1930, the year the Depression first began to bite. But Knudsen’s former boss then fell to number two in 1931, and Chevrolet would outsell Ford every year—except for 1935 and 1945—until Ronald Reagan’s second term, in 1986.63
Knudsen had won. There’s a picture of him and a trio of Chevrolet engineers admiring their new model for 1932. For once Knudsen isn’t wearing his hat. Instead, he is smiling a smile of quiet satisfaction. He had ample reason to be pleased, not just because the ’32 Chevy, except for its lack of automatic transmission and air-conditioning and a V-8 engine, was pretty much the same automobile Americans would be making and driving for the next forty years, but also because Knudsen had built product improvement into the manufacturing process in a way that could go on ad infinitum. It was the bedrock of customer loyalty.
Sloan was blunt about why. “We want to make you dissatisfied with your current car,” he told Chevy customers, “so you will buy a new one.” But there was a more profound truth underlying flexible mass production. Companies could now change how a product was made—or even introduce a new product—in rapid response to either changing market demands or to new technology, without breaking a single stride. This was true whether they were making cars or radios—or later, tanks, planes, and radar sets.
In 1936 Sloan’s GM was selling more cars than it had before the Depression. In 1937 it was selling more. By giving Knudsen and Chevrolet their lead, he had made General Motors the largest industrial corporation in the world.
Freedom's Forge: How American Business Produced Victory in World War II Page 4