The Chip Industry's Role in East Asian Catch-Up Growth
The strange nexus of Texan and Californian chip businesses, Japanese Reparations, Asian Dictators, and East Asians leaving the farm to assemble chips
I’m continuing my semiconductor series. To see part I, click here:
How it worked in the Olden Days
In the late 1800s and early 1900s, electronics weren’t really a thing. Instead, firms employed "calculators", people who tabulated logarithms, sines, cosines, and recorded decimal reciprocals manually.
The landscape began to change in the 1930s, with American, Japanese, and European entrepreneurs and governments investing in mechanical computing machines, primarily for defense purposes. These early efforts were later replaced by the development of electronic computers using vacuum tubes.
A significant milestone came in 1945 when the University of Pennsylvania created the Electronic Numerical Integrator and Computer (ENIAC), a massive machine designed to calculate artillery trajectories. This behemoth, powered by 18,000 vacuum tubes, occupied entire rooms.
How Greed Breeds Advancement
Bell Labs (now AT&T) in 1947, outdid universities when they used germanium transistors as a replacement for vacuum tubes. William Shockley, a researcher at Bell Labs, wanted to become filthy rich, so he made Shockley Semiconductor in Mountain View, California to mass produce and sell transistors. That same year, John Bardeen, Walter Brattain, and William Shockley invented the first transistor at Bell Labs.
The subsequent mass production of semiconductors, facilitated by techniques such as “Photolithography” which makes tiny, exact design patterns on silicon. This technique is crucial for creating the intricate parts in electronic devices.
In 1958, Jack Kilby, an engineer at Texas Instruments, found a way to reduce the number of wires needed to string different transistors together. He invented the way to slap multiple transistors on a slab of silicon or germanium - this is the integrated circuit or chip.
How Stealing Leads to Advancement
As technology was being sold and exported, many countries became really good at obtaining chips and then reverse engineering them to make better chips for their strategic uses. One of the best “steal, reverse engineer, & improve” nations in the 1950s was the Soviet Union.
Post WW2, the Soviet Union had an economic boom. The average Soviet went from poorer than a Colombian or Portuguese in the mid 1940s, to becoming richer than both of them by 1965.
If I was talking in the 1800s, the best “steal, reverse engineer, & improve” nation would be the the Northeast United States, stealing IP from the United Kingdom to make automated textile manufacturing.
Anyways, thanks to theft and improvements, in 1959, Soviet Union launched Sputnik and then in 1963, they launched a cosmonaut to space.
New Techniques create New Companies
Sputnik scared the US government. Also in 1959, Egyptian-American, Mohamed Atalla and South Korean-American, Dawon Kahng, at Bell labs made the metal-oxide semiconductor field effect transistor (MOSFET). This invention laid down the foundation of mass production of transistors and integrated circuits.
Eight people in Bell Labs left and received seed funding from a venture capitalist investor, Arthur Rock, to make Fairfield Semiconductor.
The new firm used MOSFET to make chips smaller, allowing more components to be stored on a chip while increasing chip performance, which is basically “tech magic”.
In 1961, Fairchild Semiconductor made a silicon chip with 4 transistors. Then the firm increasingly found a way to stuff more transistors in a chip. Gordon Moore, the founder observed in 1965 that the number of transistors on a semiconductor chip would double approximately every two years, leading to an exponential increase in computing power and a decrease in cost per transistor. This is known as Moore’s Law. As long as businesses can continue to shrink transistors and decrease the cost to manufacture them per unit, then technology will continue to advance.
Government Purchases, a Subtle Subsidy
NASA, the airforce, and the military wanted to beat the Soviets so Fairfield and Texas Instruments competed to provide chips for them.
Eventually, after getting more customers, the goal for Texas Instruments was improving the manufacturing yield - share of transistors that worked. Chinese-American, Morris Chang, figured out how to increase the yield on production line.
By the mid 60s, these chips were then making the Apollo spacecraft, missiles, satellites, sonar, and torpedoes. As in, chips were primarily used for military and space applications. There wasn’t a big consumer market.
The goal for a semiconductor firm was R&D to put more transistors in a chip —> sell chips to the government —> reinvest money into R&D —> rinse & repeat.
The U.S. Defense department in the 1960s shockingly cut defense spending (in 1962 defense spending was $54.65B, then the US cut spending for years until 1966 when it shot up again) and negotiated down prices for chips. Since the government was being stingy paying for chips for a few years, this gave the perfect impetus for the industry to find consumer applications for chips - the personal computer industry.
As firms like Fairfield were transitioning to consumer goods, in order to market to a mass of consumers, costs had to be low. So the company wanted to look for a cheap labor pool for assembly. So the question was where?
Which place has the best cheap labor?
If you were a country that was aligned with the Soviet Union, Cuba, or China, that area was a “no go area”. American President, Harry Truman, signed the “Export Control Act of 1949” which limited export of strategic materials to the Soviet Union and their allies (China, Mongolia, Poland, Cuba, Czechoslovakia, Hungary, East Germany, Romania).
Basically most semiconductor investment was in Asia. Specifically Israel and non-communist East Asia. Why?
Why Israel?
For Israel, its because Israel was loaded with high skilled and highly educated Jewish scientists, engineers, and etc. from Europe. IBM, Broadcom, and Intel started investing in Israel since the late 60s and 70s. During the 60s, American wages were 2x to 3x Israeli wages.
Why East Asia?
As for East Asia, they already had an industrialized power already investing in the region. Japan, the world’s first non-white industrialized nation, was already using places like British Hong Kong, Taiwan, Singapore, Thailand, Philippines, and South Korea for cheap labor. Japan made infrastructure investments there and chip & electronics firms like Hitachi, NEC, Toshiba, Fujitsu, and Mitsubishi Electric outsourced assembly to these regions and they quickly developed chip assembly clusters.
Even though Africa, India, and Latin America had abundance of cheap labor, they lacked the decent infrastructure and talent. Japan already developed talent in East Asia, creating a “brand” that Asian labor is the optimal labor for cheap costs. So it was easier for American and European chip firms to use existing talent in East & Southeast Asia rather than use other regions.
Back then British Hong Kong & Singapore were poor (by American standards) so Fairfield invested there, then Texas Instruments, Motorola, and other firms followed. In 1966, the Average American made $38K a year in 2022 dollars(using an inflation adjustment calculator). Meanwhile, Hong Kong Chinese made $6.4K a year in 2022 dollars, Singaporeans made $4.6K a year, Malaysians made $3100 a year, Taiwanese $2201 a year, Philippines $1894 a year, and South Korea had incomes lower than many African countries at $1263 a year in 2022 dollars.. In other words, American wages were 6x to 32x higher than these non-communist, pro-Western Asian countries, making these spots perfect places to invest.
Singapore and Hong Kong had decent infrastructure, good investment climates, and low wages - a match made in heaven. How did that happen? This synergy emerged due to various factors, including domestic investments from Singapore and Hong Kong, as well as Japan’s paying reparations for raping, colonizing/occupying, and murdering 10 million Asians. Japan provided infrastructure loans and technical aid after the San Francisco Peace Treaty.
Japanese Reparations
A non-exhaustive list of Japanese reparations:
$550M reparations to Philippines
$5M to rape victims across Asia (especially Korea)
Divestiture of $24B in Japanese assets across Asia
$220M reparations in Indonesia
Japanese Bilateral aid development
Secondly, besides throwing money to address war crimes, Japan was making infrastructure and providing technical assistance to Asian countries to make roads, dams, communication networks, rivers, and energy projects in bilateral assistance.
Throughout the 60s and 70s, Japanese taxpayer money were building railways, roads and dams in South Korea and power stations, highways, and water systems in Indonesia. This extended to Hong Kong, Singapore, Malaysia, Thailand and Philippines as well.
Japan’s Multilateral Bank - the ADB
Thirdly, Japan created the Asian Development Bank in order to foster its Pan-Asian vision of regional cooperation. (It was headquartered in Manilla, Philippines, in order to get by in from other Asian countries, that Japan was no longer interested in Japanese Supremacy). Japan’s Asian Development Bank was able to raise funds in Europe, Japan, and North America with a AAA bond rating, and then deploy below-market loans to poor Asian countries.
$5M loan to Thailand to fund its industrial development bank (1968)
South Korean highways (1968)
Water supplies for Malaysia (1968)
$10M industrial plant in Taiwan (1968)
Tajum Irrigation Project for Java, Indonesia (1969)
Japanese reparations, Japanese bilateral aid, Japanese lending via its development bank, combined with the poor East Asian nations’ local investments, transformed these non-Communist, East Asian countries into locales 1st/2nd World infrastructure in an area with 3rd world wages. There was the trinity of low transportation costs, minimal labor regulations, and low wages. I can only imagine how much an American firm was frothing off the mouth to outsource to East Asia. East and Southeast Asian investment was the steal of the century in 1960-1980s.
When Asian Socialist states adopt more pragmatism
During the 1960s to 1980s, the communist bloc, including China, missed out on these opportunities. Foreign investment was non-existent under Mao Zedong's rule, and only took off during Deng Xiaoping’s reform era. In addition, Vietnam only saw significant FDI after the Doi Moi reforms in the mid 1980s. Other socialist/communist states such as Mongolia, Laos, Cambodia, and Burma similarly lagged behind in attracting FDI until adopting partial neoliberal reforms in the late 80s/early 90s.
The African -East Asian Divergence:
These factors started the African-East Asian divergence. Due to a combo of:
Regional Power Influence: In 1955, South Africa & Japan had roughly equal incomes (Or Rather the White South Africans had European living standards and black Africans lived in squalor).
Japan, the only industrialized power in Asia during the 1960s, brought reparations and infrastructure investment in the region. That led to Japanese firms investment across East Asia, which led to American and European firms investing as well - this led to semiconductor investments.
Meanwhile South Africa, the only industrialized power in Africa during the 1960s, was murdering black Africans (like Sharpeville) in its borders and destabilizing the Southern African region to prevent nearby colonies from governing properly or receiving independence.Developing Financing Differences: Japan’s Asian Development Bank was far more capitalized than the African Development Bank. Japan’s ADB had $1B at the start. Africa’s Development bank had $100M.
East Asians grew farm yields beyond subsistence in the 50s,60s, & 70s while most African yields are still at subsistence today.
East Asian industrial policy & East Asian entrepreneurs had much better execution than African industrial policy & entrepreneurs. There is no equivalent to a Hitachi, Sony, Samsung, TSMC, or NEC in the Africa continent in terms of revenue.
East Asian countries purposely lowered their currency to boost exports. This is when a country’s central bank “financially sterilizes” the economy to keep a currency weak by selling its own currency to buy foreign currency. This weakens the local currency, and strengthens the foreign currency. East Asian countries have weakened their currency on purpose to undercut the west on exports.
Meanwhile African countries have been and still do the exact opposite. African countries purposely strengthen their currency to make food/medicine/fuel imports relatively cheap. If the official currency rate is stronger than the black market or parallel currency rate, then you know that country is/has been artificially strengthening the currency.
These five factors contributed to the great African-East Asian divergence. Africa and East Asia were roughly equally poor from 1950-1965. But by, 1965 to 1970, the divergence started and expanded and then East Asia took off.
Labor Union Suppression in East Asia
East Asia was great for Western and Japanese multinational semiconductor firms. Singapore, Taiwan, and etc. routinely crushed their unions and labor organizers, so productivity growth exceeded wage growth in these states for some time. As Charles E. Sporck from Fairchild Semiconductor said “We had union problems in Silicon Valley. We Never had any union problems in the East.”
Tens of thousands of landless East Asian peasant farmers transitioned to being packagers, assemblers, and testers for chips.
Fairfield, Hitachi, Sony, and others would make silicon wafers but ship semiconductors to Hong Kong/Malaysia/Singapore/South Korea/Taiwan/Philippines for final assembly. These workers were excellent in production quality, while assembly lines in America and Japan were too expensive.
By the late 60s and early 70s, the poor East Asian and Southeast Asian countries were assembling chips for American firms which were being used to kill Vietnamese people in the Vietnam War.
There was a weird nexus of Texan and California chipmakers, Asian Dictators, and East Asian workers assembling semiconductors for cheap wages. (Cheap by American standards, but better than farming by these Asian nations’ standards). Urban cities which used to be breeding ground for Maoist funded communist revolutions were now places of diligent Asian workers.
Conclusion
The American semiconductor industry used to have its main customer being the US military. But as the US military briefly cut spending, American chip firms needed to pivot to the consumer market. Japan was already using lesser developed East Asian countries for chip assembly & packaging, due to their proximity and Japanese investments to make these places worth investing in. The result was American firms and European firms used these East Asian nations for assembly as well. This helped create this triad chip supply chain between America, Europe, and East Asia. This also contributed to East Asia, which was as poor as Sub-Saharan Africa in the 50s, catch up to the West.
Final Note
It’s not entirely true that capitalists want to invest where labor is cheapest. That statement is incomplete. Capitalists want to invest where labor, land, and transport costs are cheap and where capital and energy are abundant (when I say capital, I mean capital as in “stuff to make more stuff”).
If capitalists wanted to invest only in cheap labor, Sub-Saharan Africa would have been a destination. But because their transportation networks were awful, East Asia makes sense. This is is still true today. Singapore, a small island with 5.5M people has 4x the net-FDI than all of Sub-Saharan Africa combined. Why? Singapore has expensive labor costs, but is abundant in capital, has abundant energy, and has good transportation.
Like, comment, restack, share and subscribe!
The Asian Development Bank History Book and Chip Wars were a great help writing this article
Thank you for this great continuation of the chip topic.
Japan's development assistance was not motivated by benevolence. The infrastructure investments they made across Asia was for two reasons.
First - they had an over-abundance of construction capacity after the post war economic boom. They needed to push this capacity onto other countries. This also motivated China's BRI projects in the late 2008s and 2010s.
Second - as you mentioned in the 80s manufacturing was based around regional supply chains. In order to make Japanese manufacturing competitive with North America and the EU (which was expanding into Southern and Eastern Europe) they invest in their neighbours. To enable these manufacturing supply chains in Asia Japan also had to build the infrastructure.
America also invested its overcapacity into Latin America. First in the early 20th century to push out British. The second after WW2 to compete with communism. Both times the Latin Americans wasted this opportunity to industrialise. They defaulted on their debts both times. Mexico might actually make it this time around despite all their challenges.