For the past 40 years of the history of personal computers, Moore’s Law — the notion that as processing power doubles every two years, computer prices fall accordingly — has rendered the market for PCs one of the most competitive in the American economy. That is, until now.

After Microsoft jacked up the prices of its Surface PC lineup by $250 to $500, Apple has followed suit, with the price of MacBook Air boosted by $200 and the MacBook Pro by $300. In the first five months of this year, the consumer price index for PCs and related devices, such as hard drives and tablets, has risen 6%, putting it on pace for a double-digit annual inflation rate.

The culprit? The bipartisan CHIPS and Science Act of 2022, which sought to solve the national security crisis of internationally vulnerable semiconductor supply chains by creating a new conundrum here at home.

Nobody denies that the status quo of chip manufacturing — in which the majority of the planet’s chips were produced in Taiwan — had become increasingly untenable in light of the increased aggression of the Chinese Communist Party and the explosive demand for semiconductors by America’s ascendant artificial intelligence industry. Rather than reduce the regulatory burdens that prevented the onshoring of semiconductor manufacturing, then-President Joe Biden used the CHIPS Act to throw money at the problem, providing roughly $52 billion in direct subsidies for semiconductor manufacturing and research, plus a 25% investment tax credit that could push the total cost north of $70 billion. Nearly four years after the bill was signed into law, the share of global semiconductor manufacturing conducted on U.S. soil rose from 10% to 12% — hardly a stunning success considering the taxpayer price tag.

Further, by picking winners in the tech industry, Uncle Sam inadvertently crowded out a newly-created class of losers: personal computing. Some tech giants, backed by government cash, have been able to devour a dominant share of computer memory production capacity, colloquially known as RAM. The small cities of new data centers springing up across the country, and Claude’s ability to schedule, troubleshoot, and draft a customized strategy for individual Americans requires terabytes on petabytes on exabytes of RAM. But personal computing hasn’t gotten any less important, even as AI’s market cap has soared.

Subsidies do not fix supply constraints; they redistribute capital, often toward politically favored firms, without lowering the underlying cost of investment across the economy. The CHIPS Act failed to ameliorate the tax burdens, regulatory barriers, and permitting delays that obstruct both new semiconductor manufacturing and new memory development. The demand for RAM was fueled by the CHIPS Act, while little was done to allow for new supply.

RAM production remains concentrated among three companies: Samsung, SK Hynix, and U.S.-based Micron, and together they control roughly 95% of global supply. The CHIPS Act directed more than $10 billion toward firms like these, but building a cutting-edge semiconductor fabrication plant takes five to seven years under ideal conditions, and often longer in the United States.

Environmental reviews, permitting delays, labor mandates, and compliance requirements all increase the cost of capital and extend construction timelines. While companies in Taiwan or South Korea can move from groundbreaking to production faster, projects in the U.S. frequently face years of delay. Even flagship investments, such as Intel’s Ohio facilities, have been slowed by funding uncertainty and regulatory complexity.

Coupled with sky-high private valuation and publicly-funded subsidies, chipmakers have rationally redirected limited capacity toward high-margin memory products for hyperscale data centers operated by companies such as Microsoft, Amazon, Google, and Meta. AI workloads now consume a majority of advanced DRAM output, leaving the PC industry, automakers, and defense systems scrambling for a smaller supply of more expensive memory.

Supporters of the CHIPS Act argue that it laid the groundwork for long-term domestic capacity. But even if that proves true years from now, it does nothing to address the immediate shortages and price spikes consumers face today. Worse, by distorting capital allocation and favoring incumbents, the policy may have slowed more efficient private investment.

You don’t get more supply by subsidizing production; you get it by making production easier, faster, and cheaper. That means lowering taxes on investment, streamlining permitting, cutting regulatory red tape, and allowing market competition to reward speed and scale.

If policymakers were serious about reshoring semiconductor manufacturing, they would focus on reducing the cost of building in America, not writing larger checks to politically connected firms. Instead, Washington chose industrial policy over supply-side reform.

The result is exactly what basic economics would predict: less supply than promised, higher prices than expected, and continued dependence on foreign production.

If America wants more chips, it needs less Washington.

***

Tim Doescher is the executive director at Unleash Prosperity and a former White House official in the George W. Bush administration.

Menu

The Real Reason Your MacBook Is More Expensive