Ten months later, Noyce proposed to Bowers on the terrace of a restaurant in Athens. They had talked about Noyce’s affair, and Bowers was confident it would not happen again. She had spent Sunday dinner with his parents and felt welcome. Even the two Noyce children whom she met seemed grudgingly accepting of her. And yet, Noyce’s proposal—a month before his divorce was finalized—surprised Bowers. She had briefly been married once before and would have felt happy just to live with Noyce. But he thought it would be best for his children if he were married, rather than living in proverbial sin, and after she thought about it a bit, she decided he was right.
A few days after Bowers tearfully agreed to marry Noyce, she offered him a proposal of her own. He wanted her to get a pilot’s license so they could fly together. She said that she would do it if he quit smoking. And she made a pledge to herself: when it was clear to her that he truly had kicked the habit for good, she would change her last name to Noyce.
He promised that he would try to give up cigarettes. Bowers began flying lessons. Noyce puffed on. When Bowers became a licensed pilot, he was still a smoker. He knew he ought to quit. When he skied a particularly challenging run or tried to speed up the laps he swam every morning, he felt his lungs straining. The surgeon general had been slapping warning labels on his cigarette packs since 1965. Noyce tried buying low tar or low nicotine cigarettes, but he would always break off their filters or just smoke more. A friend who worked in a hospital pointedly showed Noyce a piece of a diseased lung that had belonged to a smoker. “That’s amazing,” Noyce said, staring at the holes where the alveoli should have been. But still he smoked his Camels. Bowers forbid him to smoke in the house. He moved to the porch. She tried what she called “the Chinese water torture” approach to the topic, gently bringing it up time and again. No effect. She tried yelling. No better. She arranged for him to meet with a Stanford doctor who was an expert in nicotine addiction. Noyce left the meeting with an intellectual appreciation for the complexity of his addiction and a desire for a cigarette. Bowers even considered refusing to fly with Noyce—after all, he had not held up his end of their deal—but she enjoyed it too much to give it up, and besides, she knew her boycott almost certainly would have no effect on his smoking.
The closest Noyce came to quitting was a few years after his marriage, when a group of his ski buddies issued what Bowers called “a man’s challenge”: “If you quit, we’ll pay for you to ski with us in the Bugaboos. But we don’t think you can do it.” Noyce immediately stubbed out his cigarette. He did not pick up another one for nearly a year. Just as Bowers began to think about changing her name, Noyce bummed a cigarette in a parking lot and immediately returned to his smokestack ways. He had to pay his own way to the Bugaboos that year.
Smoking somehow steadied Noyce. At times when he was forced to sit still—on the telephone, in a meeting, on a commercial jetliner—the rituals of smoking (banging the pack on a table, selecting a cigarette, putting it between his lips, lighting it with one hand cupped around the flame, inhaling deeply) gave him something to do. When he was worried or angry, he smoked almost constantly, lighting a new cigarette from the stub of the one still in his mouth. At other times, he reveled in the sensual pleasures of smoking. He could recall the advertising slogan for Lucky Strike cigarettes—“so round, so firm, so fully packed”—decades after it had been retired. Several people, including one of Noyce’s daughters, have conjectured that somewhere deep within himself, Noyce thought that he was immortal. From the moment he leapt with his glider from the roof of a Grinnell barn, Noyce never stopped at the edge of a precipice but instead ran, at full speed, right over the edge and into the unknown. This had been his approach with ideas, with companies, with skiing, with driving, with women, and with inventing. How on earth could something as mundane and trivial as a cigarette kill someone like him?2
WHEN BETTY NOYCE FIRST LEARNED of Bob’s plans to marry Ann Bowers, she reacted with her characteristic biting wit. She told him that this move would make her the largest “single” stockholder in Intel. To be sure, Wall Street watched Betty Noyce’s actions quite closely. In 1976, when the stock of nearly every other semiconductor company was rising, Intel’s temporarily fell on rumors that she planned to sell 100,000 shares.3
Once permanently settled in Maine and out from Bob’s shadow, Betty Noyce put her half of the Intel money to good use. She developed a dynamic, entrepreneurial, low-profile approach to what she called “catalytic philanthropy,” starting a bank, rescuing a floundering bakery, underwriting one-third of the cost of a children’s hospital, restoring several buildings, and launching a market in downtown Portland. When she wanted to donate $1 million to Maine Public Broadcasting, she did not simply write a check. She built five houses (thereby employing dozens of people) and donated the money from the homes’ sales. At the time of her death in 1996, the $75 million she had given to the people of her adopted state ranked her among the most generous philanthropists in Maine’s history.4
Bob and Betty Noyce’s divorce was final on August 12, 1975. Noyce and Ann Bowers were married 11 weeks later, on November 27, Thanksgiving Day and Bowers’s 38th birthday, at Noyce’s home. Until that time, they had kept their relationship a secret from nearly everyone at work. Their families, who had no idea Noyce and Bowers were even engaged, thought they were simply coming to Thanksgiving dinner on the day of the wedding. After the meal, Bowers opened two boxes. Each held a bouquet, one for her and another for her five-year-old niece, who would serve as a flower girl.
“Why do you think we have these?” Bowers asked the little girl.
The child’s eyes grew wide. “Because someone is getting married?”
Noyce and Bowers looked at each other. “That would be us,” Noyce said.
The dinner guests began shrieking. Harriet Noyce started to cry because Bob’s father, who had suffered a debilitating stroke, was unable to perform the ceremony. The minister, who had hidden himself in a closet, stepped forward to marry the couple in a ceremony from which Bowers had excised every reference to God. “Bob agreed to that. Neither of us could decide about God,” Bowers says. “I remember Bob saying, ‘Some people who believe in God are good, and some people who believe in God are not good. So where does that leave you?’ He had [also] looked around and decided that religion is responsible for a lot of trouble in the world.” Noyce, always pushing against the limits of accepted knowledge, told Bowers that what bothered him most about organized religions was that “people don’t think in churches.”
But for all his insistence that “an ethical life is what really matters,” Noyce did not want a secular wedding. When his brother Gaylord could not fly in to perform the service, Noyce said, “Well, we’ll have to find a minister here.” There was no thought of a civil ceremony.
On the Monday after her wedding, Bowers went to see Andy Grove. “Bob and I got married on Thursday,” she said. “I think I should resign.” Grove disagreed. She had been at Intel almost since its founding. Anyone would be able to see that she had not gotten her job through nepotism. “Anyone can see that now,” Bowers corrected him. “But in a year or two, this will start to look very bad for Intel.” Still Grove protested. “Let Bob resign,” he suggested. He was not at all joking.5
In the end, Bowers agreed that she would stay at Intel long enough to hire and train her replacement. Finding someone took until April. In May, just when Bowers was ready to leave, Gordon Moore asked her not to go—at least not yet—because a union was attempting to organize Intel’s workers.
In the first week of May, the International Union of Operating Engineers Local 39 filed a petition seeking a representation election for the janitorial staff at Intel’s Santa Clara headquarters. The National Labor Relations Board soon dismissed the petition with an explanation that “the union has failed to demonstrate to the NLRB that a sufficient number of [Intel] employees in an appropriate bargaining unit desired representation by this union.” But Intel could not breathe easy. On June 4, the
International Brotherhood of Teamsters, Local #296, filed a petition seeking representation of certain Intel “warehousemen, drivers, shipping and receiving clerks, TWX operators, PBX operators, and mail service personnel.”6
Moore’s letter announcing the Teamsters’ efforts read, in part: “In case you have any questions regarding the position of Intel’s management with respect to union representation of any of our employees, I would like to reiterate it. Union representation of any group of employees would have a serious impact on all of Intel’s operations. We wish to continue to seek development and growth opportunities for each employee through the growth of Intel. This goal can best be achieved if each individual continues to be free to communicate directly with anyone else in the company to solve problems or to seek help or information.” This is a muted version of Moore’s strong anti-union sentiments, which were shared by nearly all of Intel’s senior management, including Noyce.
The Intel team was not alone in their beliefs. In 1973, WEMA, an industry association to which Intel belonged and in which Noyce actively participated, offered a two-day seminar for “companies that are non-union and wish to remain so.” Led by an attorney who specialized in labor law, the seminar featured a simulated union organizing drive, so participants could practice making decisions in realistic scenarios. WEMA (which changed its name to the American Electronics Association [AEA] in 1977) also provided legal aid to companies facing union drives and furthermore served as a highly efficient clearinghouse for information about union activity throughout the electronics industry. As one union organizer explained, “Whenever organizers passed [out] leaflets in one plant, a copy of the leaflet would be on the desk of every human resources director in the Valley within two or three days.”7
Conventional wisdom within the semiconductor industry held that no matter how rich a company’s wages and benefits package, it would cost 25 percent more to operate the business with a union in house. A strike in April 1968, by 5,000 workers at Ampex, Lenkurt Electric, and Dalmo-Victor—the only three unionized electronics (not semiconductor) companies in Silicon Valley—reinforced the prevailing bias against unions. The work stoppage stretched more than a week. Had this lacuna hit a company in the fast-moving semiconductor industry, it would have put the firm at a significant competitive disadvantage.8
Intel had long employed many of the same preemptive personnel tactics that had kept unions at bay at Fairchild. Hourly workers at Intel had health care, dental care, paid vacations, and sick leave. They could buy Intel stock at a reduced price. The company also held meetings each month at which supervisors answered questions from employees, either asked in person or submitted anonymously in writing. Such practices, common in Silicon Valley semiconductor firms since the early Fairchild days, were widely encouraged by the AEA, one of whose representatives warned, “The way to thwart unions is to make them unnecessary. And the way you do that is to think as though you really had a union in the plant.”9
Intel had union-thwarting personnel policies on the books, but policy and practice diverged in 1975 and 1976, when the industry pulled out of the 1974 downturn and Intel resumed its dramatic growth. By the second quarter of 1976, when the unionization efforts began, demand for Intel’s latest-generation memory devices exceeded production capacity. This led the company to open a new fab near Portland, Oregon and a testing facility in Santa Cruz, California, while also beginning construction on an assembly plant in Barbados, West Indies. At the same time, Intel was hiring at a furious rate. More than 1,250 new employees joined the company in the last half of 1975—a rate better than 200 each month. This breakneck growth continued into 1976, which saw another 2,700 employees join the company.
In the midst of all this activity, everyone at Intel felt pressured to put in ever-longer hours, and even Andy Grove did not notice that a foreman at one of Intel’s facilities had given every employee identical merit increases. This was anathema. Intel always awarded performance-based increases. The company said this policy rewarded individual effort; the labor movement claimed its primary effect was to discourage collective thinking of any sort among workers.
In any case, Intel’s failure to follow its own policies led someone at the company either to contact the Teamsters or to listen carefully when approached by them. “I felt we brought the unionization problem on ourselves,” Bowers explains. “If we had followed our own policies, it never would have happened.”
The timing of Intel’s misstep coincided with a spike in union-organizing activity throughout Silicon Valley. In 1974, the United Electrical Workers (UE) created an organizing committee specifically to target the Silicon Valley labor force. Many in the labor movement thought that workers might be more open to collective bargaining after the massive 1974 layoffs, especially if someone pointed out to them that with every passing year, more semiconductor production jobs moved offshore.
A worker’s sense of insecurity may have been heightened by the increasingly austere and antiseptic nature of semiconductor production work. Even Intel’s head of manufacturing, Gene Flath, admits that semiconductor fabs had become increasingly “scary places to work” by the mid-1970s. Intel’s Fab 3, completed at Livermore, California, in April 1973, was twice the size of its predecessors and specifically designed to be “a big 1103 machine.” Every surface in its $2 million, 10,000-square-foot clean room gleamed, and hanging on many walls were huge warning signs reading CAUTION or ACID. Here Intel employees, who had once worn lab coats that they embroidered and shortened into minidresses, were required, for the first time, to don head-to-toe “bunny suits” with hats, goggles, long sleeves, pants, and gloves. They were forbidden to wear makeup. The suits and scrubbed faces were designed less to protect the workers than to keep the manufacturing climate uncontaminated—even a wayward flake of human skin or mascara caused problems.10
In part because experienced fab technicians thought the bunny suits were unnecessary, Fab 3 was deliberately staffed with people with little experience in the industry. Recalls one early employee, hired at age 16 for $2.30 per hour: “We were just petrified because this was such a big job, such an important job, and it was so scary in the clean room. Everything was crammed so closely together. The aisles were very narrow. … Everything was under yellow lights, and we were petrified.”11
Intel found that a number of its new employees were quitting after only a few days on the job because, as Flath put it, “they were just so nervous they’d go home and shake at night because they [hadn’t known] what they were going to get into.” The strangeness of the clean room that terrified workers gave union organizers hope. Indeed, the most significant unionizing effort Intel faced was at Fab 3 the late 1970s.12
But that effort failed, too, as did every other attempt to unionize the semiconductor industry in the late 1970s and early 1980s. Union organizers in these years were swimming against the tide. Between 1970 and 1988, the percentage of California workers represented by a union dropped from 36 percent to 22 percent. By the end of the 1970s, nearly three-quarters of semiconductor production workers were women and almost half were members of minority ethnic groups (mostly Hispanic or Asian). Neither women nor minorities had traditionally joined unions in large numbers. Moreover, as soon as the industry’s fortunes improved, this workforce proved so mobile—in 1979, turnover rates among Silicon Valley assembly workers topped 50 percent annually—that it was difficult to organize.13
A month after he announced the Teamsters’ request to hold an election at Intel, Moore informed employees that the union had withdrawn its petition, presumably because its leaders decided that Intel workers would not support a union at the company. “Our non-union status is the direct result of your choice as Intel employees to have open and free communication and to solve our problems together, without outside intervention,” Moore reminded his employees. “We want to keep it that way.”14
A few weeks after the union scare passed, Ann Bowers quietly left Intel. She soon established a human-resources consulting business. S
he also co-founded the California Electronics Association—an organization that helped small electronics companies to develop personnel policies, find insurance, contract with credit unions, and train their employees.
Noyce, meanwhile, continued to spend a good bit of his time at the company. Even Gordon Moore says Andy Grove’s comment about Noyce “practically disappearing” is “too strong.” Noyce chaired board meetings every month, work that required regular check-ins with Moore and Grove. He attended executive staff meetings whenever he was in town. He maintained an office at Intel’s new headquarters building in which all offices were cubicles. He in short order found himself pulled into meetings at the company so often that he also commandeered an office at Bowers’s consulting firm “so I can get something done.” (He soon missed Intel too much, however, and returned to his cubicle there.) He attended Intel sales meetings and plant openings. He stood as the company’s spokesman at conferences, particularly those having to do with the microprocessor or automotive electronics. He built customer relationships by speaking to General Electric managers on “new electronics” and to Monsanto executives on “managing innovation.” He made many presentations to analysts, including an important talk to the New York Society of Security Analysts on Intel’s tenth anniversary. Noyce also continued the high-level negotiations with customers that had been his specialty as CEO.15
Intel employees who encountered Noyce in the halls during these years found their interactions with him as stimulating as ever. “I remember I ran into Noyce [in about 1976],” recalls one employee, who at the time was working on a special memory device (called an EPROM) that was one of Intel’s best sellers. “I said to Noyce, ‘I’ll bet you can’t sleep at night knowing that [I am] in charge of the [EPROM], which is a third of your profit.’ He looked at me and he said, ‘Yes, I can; I can’t think of a better guy to run the thing.’ And that made me feel good. I said, ‘Yes, maybe he’s right.’”16
The Man Behind the Microchip Page 36