David, one of my fellow college students in the early to mid 1980's, was a unique individual, a free spirit, outgoing and outspoken. He liked to have fun, and he liked other people, and everyone I knew who knew him liked him. David worked his way through his first two and a half years of college looking through a microscope. He worked for Motorola quality control in one of their semiconductor fabrication plants (fabs) finding flaws in silicon wafers. Although it has been over three decades, so my memory could be faulty, from what I remember David telling me, he could inspect as many as ten thousand wafers during an eight-hour shift. I distinctly remember him complaining many times that his job at Motorola was ruining his eyesight. On the positive side, Motorola paid for most of David's university education, and it paid for a nice place for him to live with a fellow Motorola employee named Russ. Motorola even paid for David's 140 mph plus Honda Interceptor. Once, he told me he had been pulled over by the police for speeding, and the first thing the officer said was, "Why'd you stop? We can't catch these bikes. Most people on these things just lose us."
I believe David's worsening eyesight may have been one reason that as a junior he applied for a scholarship from the air force. He received the scholarship in exchange for a four-year commitment after graduation. This seemed to me like a form of indentured servitude, and I worried that four years under an authoritarian organization would destroy the personality that we all liked so much, that it would fundamentally change David for the worse. Our jobs shape us--sometimes in subtle ways, sometimes in not-so-subtle ways. Sometimes they help to transform us into the people we are meant to be. Sometimes they help to transform us into the people we are not meant to be. As it turns out, David graduated, I went on to graduate school, and I never saw him again. I often wonder who he has become over the last thirty-something years.
Fast forward two decades to the mid 2000's and another friend of mine named Randy. Randy supported his wife and two kids on the salary he earned as an engineer working for Advanced Micro Devices at one of the last AMD fabs left in the US. If my memory is correct, there were two AMD fabs still in the US at that time. As part of his job, Randy spent a significant amount of time in China, helping to set up fabs there and train Chinese engineers. It has to be depressing to train the people you believe will soon have your job. During the whole three years I knew him, Randy was telling me how worried he was that he would soon have to move to China to continue working in his profession. Luckily for Randy, Intel bought the fab at which he worked in 2007. Wikipedia says it's still in operation today. AMD no longer owns a single fab.
At the same time Randy was worrying about his career, another casualty of the semiconductor industry's move to China, a close family friend with a masters degree in electrical engineering was laid off from his job at another well-known semiconductor manufacturer in the Northwestern part of the US. He remained unemployed for two years before he finally gave up looking for another engineering job and switched careers.
The exodus of semiconductor fabs from the US is really just one example of the push for globalization that began in the US in the 1960's. Big business pushed for it, eager for more profits, and politician's helped, eager for more campaign contributions from big business. The incentive to business becomes crystal clear when you recognize that a unionized autoworker in Detroit earns $58 an hour, including wages and benefits, and a Mexican autoworker earns $8 an hour. In the 1960's and 1970's, US voters were told that globalization would help US business penetrate foreign markets. In the 1980's, after it had became obvious that this wasn't happening, voters were told that globalization was inevitable, and if US corporations didn't participate, they would be left behind. In 1970, manufacturing accounted for 24.3% of the US GDP. By 2018, US manufacturing as a percent of GDP would fall to half of that. In contrast, China's industrial production increased by an average of over 10% per year during the same period. Manufacturing accounted for 40% of China's GDP in 2015. Do you think it's just a coincidence that most of China's political leaders are engineers?
Fast forward from the mid 2000's about decade and a half. Today, some fabs are still located in the US, but there are many, many more around the world. A large concentration are in China, Taiwan, and South Korea. Some are in Japan. A few others are scattered elsewhere around the world. Wikipedia has a list of these fabs, along with the costs of construction. Depending on when a fab was built, the cost varies anywhere from just under a billion dollars to over ten billion. Some fabs currently under construction are estimated to cost as much as twenty billion. This is a huge investment for any company to make, and a huge incentive for any company to protect its fab from outside competition.
Although I haven't talked much about the computer parts that are built with semiconductors, the computer parts manufacturing industry has followed a similar trend. Computer parts now come predominantly from Southeast Asia. And, although a few computers are still assembled in the US, many of the US assembly facilities are owned by companies based in Singapore and China.
I wrote the preceding paragraphs to give younger readers an historical perspective they may not otherwise have had. The United States was once the foremost manufacturer of semiconductors and computers in the world. Three Bell Labs scientists in New Jersey invented the first point-contact transistor, for which they received the Nobel Prize in physics in 1956. All through the 1960's, 1970's, 1980's, and 1990's, the US dominated the semiconductor industry. The PC industry was born in the US in the late 1970's, and those computers had microprocessors manufactured in the US. In 1980, 42% of the world's semiconductors were manufactured in the US. Many Americans, including many people I knew, benefited from those jobs. By 2007 the US share of world semiconductor manufacturing had fallen to only 16%. Despite the long exodus of semiconductor manufacturing from the US, it wasn't until 2015 that the rest of the world first sold more semiconductors than the US. However, those US sales included a large share of semiconductors manufactured outside the US and sold by US-based companies. Over the fourteen years from 2001 to 2015, the number of people employed in the semiconductor industry in the US fell by 38%. In 2015, of the 36 fabs planned to be built over the next two years, only 5 were in the US.
It took decades to create and build up the semiconductor and computer industries in the US, and it took the better part of two decades to tear them down, or at least to severely weaken them. This means that it is unlikely that Trump's four or possibly eight years in office will have much of an impact on where semiconductors and computers are manufactured. The implication is that, thanks to tariffs, American consumers will likely pay more for their next computers, but they will most likely not be the ones to build them and benefit from those jobs. Don't get me wrong; I like the fact that Trump is trying. I think it is long overdue. But, just as a battleship does not turn on a dime, I am sad to say that the US semiconductor and computer manufacturing industries will most likely not improve significantly during one or two terms of Trump's presidency. Even if Trump's trade war were to succeed, and even if it were to be the start of a long-term trend, many of the companies building semiconductors and computer parts in the US would still likely be foreign-owned. Much of the profits would likely still be leaving the US. However, when Trump leaves office, I see no indication that things should not go back to business-as-usual for US politicians and businesses.
Another factor affecting US jobs in a major way is automation. Trump can do nothing about this, because it has nothing to do with our national borders or with foreign competition. Fortunately, we have past examples of automation to look to, namely the industrial revolution, and specifically the replacement of farm workers with machinery. In the year 1500, between 55% and 75% of the workforce in Europe was engaged in agriculture. In 2007, there were 2.2 million farmers in the United States. This was something like 1.5% of the US workforce. Notice that we didn't have an unemployment rate in 2007 of between 53% and 73%. The reason is that people found other things to do. And there is no reason to think this won't happen every time technology changes. It may take a generation for things to adjust, but in all likelihood, adjustments will occur. The good news is that more automation is likely to make your next computer cheaper than it otherwise would have been, but not thanks to Trump.
Copyright © 2018-2019 The Cheapskate's
Guide to Computers and the Internet. All rights reserved.