CHIPS and Science Act

Based on Wikipedia: CHIPS and Science Act

In the summer of 2022, the United States government did something it hadn't done in decades: it placed a massive bet on manufacturing. Not just any manufacturing, but the tiny, impossibly complex chips that power everything from your smartphone to the missiles in military arsenals. The price tag? Nearly three hundred billion dollars.

This wasn't charity. It was fear.

The Vulnerability No One Wanted to Talk About

Here's a fact that kept defense planners awake at night: as of 2022, roughly ninety percent of the world's most advanced semiconductors—the chips smaller than twenty-eight nanometers—were manufactured in Taiwan. One island. One hundred miles from mainland China. A country that has never renounced its intention to reunify with Taiwan, by force if necessary.

Semiconductors aren't like other products. You can't just spin up a new factory when supplies run short. Building a modern chip fabrication plant, known in the industry as a "fab," takes three to five years and costs anywhere from ten to twenty billion dollars. The machinery inside is so precise that the slightest vibration can ruin an entire production run. Some fabs are built on special foundations designed to dampen seismic activity from miles away.

The chips themselves are miracles of engineering. A modern processor contains billions of transistors, each smaller than a virus. The manufacturing process involves hundreds of steps, exotic chemicals, and light with wavelengths shorter than anything found in nature. Only a handful of companies on Earth can do this work, and most of them aren't American.

This wasn't always the case. In 1990, the United States produced thirty-seven percent of the world's semiconductors. By 2020, that number had fallen to twelve percent. The factories had moved to Asia—first Japan, then South Korea, then Taiwan—following lower costs and government incentives. American companies still designed chips, but they outsourced the actual manufacturing to specialists like the Taiwan Semiconductor Manufacturing Company, known universally as TSMC.

Then the pandemic hit, and suddenly everyone understood the problem.

When the Chips Stopped Flowing

In 2021, car dealerships across America sat empty. Not because people didn't want cars—demand was surging. The problem was that automakers couldn't get chips. Modern vehicles contain thousands of semiconductors controlling everything from engine timing to seat heaters. Without them, billion-dollar assembly lines ground to a halt.

Ford lost an estimated 3.9 billion dollars in revenue. General Motors shut down eight plants. Toyota, the world's most efficient manufacturer, cut global production by forty percent. The chip shortage cost the auto industry alone an estimated two hundred billion dollars in 2021.

But cars were just the most visible symptom. Hospital equipment ran short. Video game consoles became impossible to find. The military worried about its ability to manufacture precision weapons. A single natural disaster in Taiwan, or a Chinese naval blockade, could cripple the American economy overnight.

Congress, not known for moving quickly, suddenly found motivation.

The Architect Behind the Curtain

The CHIPS and Science Act didn't appear out of nowhere. Its intellectual godfather was Keith Krach, the Under Secretary of State for Economic Growth during the Trump administration. Krach had spent his career in technology—he'd cofounded DocuSign, the digital signature company—and he understood semiconductors weren't just products. They were the foundation of modern military and economic power.

In October 2019, Krach presented a plan to Senators Chuck Schumer, a Democrat from New York, and Todd Young, a Republican from Indiana. His pitch was ambitious: transform American technology policy through what he called the Global Economic Security Strategy. The idea was to leverage one hundred fifty billion in government research and development funding into a five hundred billion dollar investment by bringing in private sector matching funds and coordinating with allied nations.

Krach even had a name for the coalition: the "Techno-Democracies-10," a group of technologically advanced democracies that would cooperate on critical technologies while excluding authoritarian states, primarily China.

The first concrete result was TSMC's announcement in May 2020 that it would build a twelve billion dollar factory in Arizona—the first advanced semiconductor fab built in America in years. Krach had brokered the deal, but he saw it as just the beginning. The TSMC announcement was meant to be a catalyst, inspiring other chipmakers to follow and giving Congress a tangible success to build upon.

Two Bills Become One

What eventually became the CHIPS and Science Act started as two separate pieces of legislation, each with its own champions and constituencies.

The first was the Endless Frontier Act, introduced in May 2020 by Schumer, Young, and Representatives Ro Khanna (a Democrat from California) and Mike Gallagher (a Republican from Wisconsin). The name was a deliberate echo of Vannevar Bush's famous 1945 report "Science: The Endless Frontier," which had laid the groundwork for America's postwar scientific dominance. The new act proposed one hundred ten billion dollars over five years for basic and advanced technology research, with special emphasis on artificial intelligence, quantum computing, advanced communications, biotechnology, and next-generation energy.

The second was the CHIPS for America Act, championed by Senators John Cornyn, a Republican from Texas, and Mark Warner, a Democrat from Virginia. This bill was more narrowly focused: fifty-two billion dollars specifically to incentivize semiconductor manufacturing in the United States. Cornyn's Texas was already home to major Samsung investments, and Warner's Virginia housed numerous technology companies dependent on reliable chip supplies.

The legislative process that combined these bills was tortuous. The Senate passed the United States Innovation and Competition Act in June 2021 with surprising bipartisan support—sixty-eight votes in favor, thirty-two against. The House passed its own version, called the America COMPETES Act, in February 2022. The two chambers then spent months reconciling their differences in what's known as a conference committee.

The final bill passed the Senate sixty-four to thirty-three in July 2022, then the House two hundred forty-three to one hundred eighty-seven. President Biden signed it into law on August 9, 2022.

Where the Money Actually Goes

The CHIPS and Science Act's headline number—two hundred eighty billion dollars—is somewhat misleading. That's the total authorization, meaning it's the maximum Congress said could be spent. Actual appropriations, the money that's actually allocated, were significantly less.

The core semiconductor provisions totaled about fifty-three billion dollars in real money. Thirty-nine billion went to manufacturing incentives—grants, loans, and loan guarantees to companies building fabs in America. Another twenty-four billion took the form of a twenty-five percent tax credit for investments in semiconductor manufacturing equipment, designed to make American factories cost-competitive with Asian alternatives.

Eleven billion was designated for research and development. Of this, eight and a half billion went to the National Institute of Standards and Technology, better known as NIST, the government agency responsible for maintaining measurement standards and conducting research on emerging technologies. Two billion funded a new entity called the National Semiconductor Technology Center, intended to serve as a hub connecting government labs, universities, and private companies.

The act also allocated two billion to the Department of Defense for military microelectronics research and five hundred million to the State Department for international semiconductor supply chain coordination. One and a half billion went to a fund promoting open Radio Access Network technology for 5G wireless networks—a technical initiative aimed at reducing dependence on Chinese equipment makers like Huawei.

But here's the catch that companies accepting these subsidies don't advertise: the law includes a ten-year ban on building advanced chip factories in China or Russia. Any company taking American money cannot simultaneously help potential adversaries develop their own semiconductor capabilities. This provision forced companies to choose sides in what some analysts have called a new Cold War—this time fought with transistors instead of tanks.

The Science Half of the Bill

The "Science" in CHIPS and Science Act often gets overlooked, but it represents the bulk of the authorized spending—one hundred seventy-four billion dollars worth of provisions for research across virtually every scientific domain.

The National Science Foundation received authorization for eighty-one billion dollars, including twenty billion for a new Directorate for Technology, Innovation, and Partnerships. This directorate was tasked with a mandate that might seem unusual for a science agency: ensuring that federally funded research actually gets deployed in the real world, with attention to social and ethical implications.

The law also invested heavily in STEM education—Science, Technology, Engineering, and Mathematics. It recommended one hundred million dollars for rural schools, tripling the Noyce Teaching Scholarships that support science teachers, and creating a three hundred million dollar STEM Teacher Corps. The explicit goal was ensuring that America could train enough engineers and scientists to staff all those new factories.

NASA received extended authorization through 2030 for the International Space Station and partial funding for the Artemis program to return humans to the Moon. Perhaps more ambitiously, the law directed NASA to establish a Moon to Mars Program Office, laying groundwork for eventual human missions to the red planet. It even included provisions for nuclear propulsion research and continued funding for the search for extraterrestrial intelligence.

The Department of Energy received authorization for new research initiatives in supercomputing, nuclear fusion, and particle physics, plus the creation of a new nonprofit organization called the Foundation for Energy Security and Innovation to leverage private philanthropy for energy research. Climate provisions authorized sixty-seven billion dollars for zero-emissions technologies, improved energy storage, and hydrogen economy development.

And in a nod to economic geography, the law directed the Department of Commerce to establish ten billion dollars worth of regional technology hubs in communities that had been left behind by previous waves of industrialization—former manufacturing centers and rural areas that the new economy had largely bypassed.

The Security State Provisions

Buried within the CHIPS Act were provisions that had nothing to do with computer chips and everything to do with a different kind of competition with China: the battle for scientific talent and research integrity.

The law imposed new restrictions on Confucius Institutes, the Chinese government-funded cultural organizations on American university campuses that critics argued doubled as platforms for surveillance and influence operations. It established new research security offices at the Department of Energy, NIST, and the National Science Foundation, tasked with preventing foreign governments from stealing federally funded research.

New disclosure requirements targeted what the law called "foreign talent recruitment programs"—arrangements in which American researchers accepted money from foreign governments, particularly China's Thousand Talents Program, which had been accused of systematically harvesting American scientific advances. Researchers applying for federal grants would now face scrutiny about any foreign government connections.

These provisions reflected years of concern, some would say paranoia, about Chinese technology acquisition. American prosecutors had brought dozens of cases against researchers accused of hiding Chinese funding or sharing sensitive technology. Universities had been forced to shut down research collaborations. The CHIPS Act codified what had previously been ad hoc enforcement into systematic policy.

The Results So Far

By March 2024, analysts estimated the CHIPS Act had catalyzed between twenty-five and fifty potential semiconductor projects, with total projected private investments between one hundred sixty and two hundred billion dollars. The projected job creation ranged from twenty-five thousand to forty-five thousand new positions.

The biggest announcements grabbed headlines. Intel committed to a twenty billion dollar expansion in Ohio, building what it called the largest semiconductor complex in American history near Columbus. TSMC's Arizona project grew from twelve to forty billion dollars, with plans for multiple advanced fabs. Samsung announced a seventeen billion dollar facility in Texas. Micron pledged up to one hundred billion over two decades for a new fab in upstate New York.

But the reality proved messier than the press releases suggested.

Projects faced delays in receiving promised grants as the Commerce Department struggled to process applications. The bureaucratic machinery designed to prevent waste and fraud slowed disbursements to a crawl. By early 2024, companies complained they were building factories on hope, unable to count on government money that remained stuck in administrative review.

Skilled workers were even scarcer than capital. A semiconductor fab requires thousands of specialized technicians—process engineers, equipment operators, cleaners who can work in sterile "cleanroom" environments. These workers don't appear overnight. Training programs take years, and American educational institutions had largely stopped producing semiconductor specialists when the industry moved offshore decades earlier.

Then came the budget battles. The CHIPS Act authorized ambitious spending levels, but Congress must appropriate money separately. In subsequent budget negotiations, lawmakers cut or limited research provisions by tens of billions of dollars. The gap between what the law promised and what companies could actually expect widened.

The Critics Have a Point

Senator Bernie Sanders voted against the CHIPS Act, one of only a few Democrats to do so. He called it a "blank check" to profitable corporations, questioning why taxpayers should subsidize companies like Intel that had spent billions on stock buybacks while closing American factories. House Republican leader Kevin McCarthy, from the opposite end of the political spectrum, used the exact same phrase.

Their critique contained uncomfortable truths. Intel's market capitalization exceeded one hundred billion dollars. Samsung was one of the largest companies on Earth. TSMC's founder, Morris Chang, had become one of the world's wealthiest people. Did these corporations really need government subsidies to build factories?

The semiconductor industry's response was straightforward: economics. Building a fab in America cost roughly thirty percent more than building one in Asia, even before considering ongoing operational costs. Without subsidies, companies would simply continue building overseas, where governments had been offering generous incentives for decades. Taiwan, South Korea, and Singapore had all used industrial policy to capture the semiconductor industry. America was just catching up.

But catching up meant playing a game other countries had already been playing. In 2022, the European Union proposed its own European Chips Act with forty-three billion euros in support. China was already spending heavily on domestic semiconductor production, despite American efforts to restrict its access to advanced manufacturing equipment. The CHIPS Act might not start a subsidy race—that race was already running—but it certainly accelerated it.

Analysts at the Brookings Institution raised concerns about the workforce development provisions, which they found vague and fragmented. The law created dozens of separate training programs spread across multiple agencies with no coherent strategy. Provisions for supporting workers in caregiver roles or addressing gender and racial gaps in the technology workforce were aspirational rather than operational. The Department of Commerce's own inspector general found the agency hadn't even planned for hiring the staff needed to implement the programs.

What China Thinks

China's government called the CHIPS Act evidence of "Cold War mentality," and from Beijing's perspective, that assessment wasn't entirely wrong.

The law was explicitly designed to contain Chinese technological advancement. The ten-year ban on investing in Chinese chip facilities meant companies had to choose: American subsidies or Chinese market access. The research security provisions targeted Chinese talent programs specifically. The entire framing of the legislation—endless references to competition, strategic autonomy, supply chain resilience—identified China as the threat requiring this massive government response.

China wasn't standing still. Despite American export controls restricting access to advanced chip-making equipment, Chinese companies like SMIC (Semiconductor Manufacturing International Corporation) continued advancing. In 2023, Huawei released a smartphone containing a seven-nanometer chip that analysts believed SMIC had manufactured, suggesting Chinese capabilities were developing faster than American policymakers assumed.

The technology competition between the United States and China has become the defining industrial policy challenge of the twenty-first century. The CHIPS Act represents America's opening move in a contest that will likely last decades.

The Bet

Industrial policy is inherently a gamble. Governments try to predict which industries will matter in the future, then steer resources toward them. Sometimes this works spectacularly—Taiwan's investment in TSMC created a company that manufactures the chips for Apple, Nvidia, AMD, and most other technology leaders. Sometimes it fails expensively—remember Solyndra, the solar panel company that received five hundred million in federal loan guarantees before going bankrupt?

The CHIPS Act assumes several things will prove true. First, that semiconductors will remain critical to economic and military power for decades to come. Given the explosion of artificial intelligence, which requires massive amounts of computing power, this seems like a safe bet. Second, that manufacturing location matters—that having fabs on American soil provides security benefits worth tens of billions of dollars. This is harder to evaluate. In a true crisis with China, would Arizona fabs really operate independently of the global supply chain for materials, chemicals, and components?

Third, and perhaps most uncertain, the law assumes the American government can effectively manage industrial policy at this scale. Historically, America has preferred market-based approaches, leaving investment decisions to private companies. The CHIPS Act represents a significant philosophical shift toward the kind of state-directed capitalism that Asian governments have practiced for decades.

Whether American institutions can adapt to this new role remains to be seen. The early implementation challenges—bureaucratic delays, workforce shortages, budget cuts—suggest the transition won't be smooth. But the alternative, watching passively as critical manufacturing capabilities concentrate in geopolitically vulnerable locations, seemed increasingly untenable.

The chips may be small, but the stakes are enormous.