The US is spending public money to make semiconductors, antibiotics, and rare earth magnets at home even though it can buy all three more cheaply from Asia. The reason is not cost. It is that a supply cutoff on any of them would cost the economy far more than the subsidy does, and the last five years supplied the proof. For us, that decision shows up as a specific kind of industrial demand: fabs, their supplier parks, and the advanced manufacturing buildings around them, concentrated in a short list of markets that includes Phoenix.
The efficiency argument is real, and it says offshore
Start with the case against doing this. Comparative advantage, the oldest idea in trade economics, says make a thing wherever it is cheapest and buy the rest. By that logic the US should not assemble much of anything. Manufacturing fell from about a quarter of US GDP in the mid-20th century to roughly 9.4 to 9.5 percent in 2025, per the Cleveland Fed and Deloitte. China now produces about 28 percent of global manufacturing value added, around $4.66 trillion, more than the next three countries combined, while the US share has slipped from about a fifth two decades ago to roughly 17 percent (IBTimes UK). That gap is not an accident. It is what happens when capital follows the lowest landed cost for forty years.
Reshoring does not reverse that on economics. Kearney's Reshoring Index stayed negative in 2025, which means the US still imported more manufactured goods from low-cost Asian countries than the reshoring story implies. Deloitte's read is blunt: bringing labor-intensive production back is expected to cost money and time even with high tariffs in place. So when someone says reshoring pays for itself, the data does not back that up. It usually does not.
So why do it anyway: three chokepoints
The case for reshoring is not that it is cheap. It is that a few supply chains are so concentrated in one country that losing access would stop the economy, and that country has shown it will pull the lever.
Rare earths are the clearest example. China controls about 90 percent of the world's rare earth refining capacity (Mitsui Global Strategic Studies Institute). Between 2023 and 2026 Beijing imposed export controls on gallium, rare earths, and at least seven other materials (Council on Foreign Relations), and the US-China Economic and Security Review Commission titled its 2025 chapter on this "Beijing's Weaponization of Supply Chains." Rare earth magnets go into F-35s, missiles, EV motors, and the robots on a modern factory floor. There is no cheap domestic substitute, which is the whole problem.
Pharmaceuticals are the same shape with higher stakes. The Council on Foreign Relations reports China controls roughly 94 percent of 6-APA, the starting material for amoxicillin, and that China and India together supply the majority of the world's active pharmaceutical ingredients (ITIF). The Congressional commission's language is that a Chinese cutoff "could be catastrophic" for US drug supply. You cannot run a hospital system on a just-in-time antibiotic supply you do not control.
Semiconductors are why Phoenix is in this article at all. The most advanced chips are made in Taiwan, about 100 miles off the Chinese coast, and everything from a car to a Patriot battery needs them. The 2021 chip shortage idled US auto plants and was a live preview of the tail risk. That is the logic behind the CHIPS Act and behind TSMC's Arizona build.
The honest part: this is subsidy, and the boom already crested
Because none of this pencils on cost, the government paid for it, and the spending shows. US construction spending on manufacturing tripled in the three years after the pandemic to roughly $250 billion, by Apollo's count, and the Census Bureau put the record near $249.8 billion in August 2024. The Reshoring Initiative counts about 247,000 manufacturing jobs announced in 2025 from reshoring and foreign direct investment, close to the 244,000 of 2024, and more than 2.3 million since 2010, with 88 percent of the 2024 jobs in higher-tech sectors.
Two cautions sit under those numbers. Many are announcements, not finished buildings, and the gap between the two is wide. And the construction wave has already crested: factory construction spending was down 22.1 percent year to date through June 2026 as CHIPS-era subsidies rolled off, and data center construction passed semiconductor fab construction in spring 2026 (Census Bureau). The reindustrialization is real, but it is lumpy, policy-driven, and front-loaded into a few sectors and a few metros.
Where it lands for us
We do not build fabs. We build and own the industrial that collects around them, and that is where reshoring reaches our pro forma. CBRE's 2026 outlook names reshoring and advanced manufacturing as structural drivers of industrial demand, forecasts industrial leasing up about 5 percent to nearly 1 billion square feet, and flags Phoenix as a market where available space may dwindle quickly. CBRE's Phoenix review ties the local strength directly to semiconductors: TSMC's Arizona commitment now totals about $265 billion and Intel's Fab 52 anchors high-wage jobs and supplier in-migration.
The built evidence is already trading. On August 10, 2026, Pivotal Manufacturing Partners and Meadow Partners announced they had acquired 53 Ten, a 157,523 square foot advanced manufacturing building on 19.6 acres in Glendale, fully leased to a Corning affiliate that serves semiconductor, aerospace, and defense customers (Business Wire). Terms were not disclosed. That is the pattern: not the fab itself, but the specialized, heavily powered, often net-leased building next to it, leased to a supplier who followed the anchor.
For an underwriter the takeaway is narrow and useful. Reshoring is not a rising tide that lifts all industrial. It concentrates in advanced manufacturing and the supplier ecosystems around a handful of anchors, Phoenix among them, and it is driven by security budgets that can be cut, not by a cost advantage that compounds. We underwrite it as real demand with a political dependency, size the building to a specific supplier tenant rather than generic distribution, and we do not pay fab-adjacent land basis on the assumption the subsidy spigot stays open. The country has decided some things have to be made here even when it is not efficient. The buildings that decision requires are the part we can own.
Sources
- Federal Reserve Bank of Cleveland, "Trends in Manufacturing in the United States" (manufacturing about 9.5 percent of GDP, 2025 Q3)
- Deloitte, "A shrinking workforce may thwart US manufacturing ambitions" (9.4 percent of GDP in Q2 2025; reshoring expected to cost money and time)
- IBTimes UK, "China Tightens Its Grip on Global Manufacturing as US Share Slides to 17%"
- Safeguard Global, "Top 10 Manufacturing Countries in 2025" (China about $4.66 trillion, 28 percent of global manufacturing value added)
- Apollo Global Management, "The Reindustrialization of America is Underway" (US manufacturing construction spending tripled to about $250 billion)
- US Census Bureau, Monthly Construction Spending (C30) releases, 2024 to 2026 (record near $249.8 billion in August 2024; down 22.1 percent year to date through June 2026)
- Kearney, Reshoring Index 2025 (index stayed negative)
- Reshoring Initiative, 2025 data report and 2024 Annual Report (about 247,000 jobs announced in 2025; 244,000 in 2024; more than 2.3 million since 2010; 88 percent of 2024 jobs in higher-tech sectors)
- Mitsui and Co. Global Strategic Studies Institute, Monthly Report, April 2026 (China about 90 percent of rare earth refining capacity)
- Council on Foreign Relations, "The Pharma Choke Point" (about 94 percent of 6-APA) and "Dismantling China's Supply-Chain Weapon" (export controls on gallium, rare earths, and other materials, 2023 to 2026)
- US-China Economic and Security Review Commission, 2025 Annual Report, "Chained to China: Beijing's Weaponization of Supply Chains"
- Information Technology and Innovation Foundation (ITIF), "How America Can Reduce Its Dependence on Chinese Active Pharmaceutical Ingredients"
- CBRE, "U.S. Real Estate Market Outlook 2026: Industrial" (reshoring and advanced manufacturing as demand drivers; leasing up about 5 percent to nearly 1 billion square feet; Phoenix flagged)
- CBRE, "Phoenix 2026 U.S. Real Estate Market Outlook Midyear Review" (TSMC Arizona total about $265 billion; Intel Fab 52)
- Business Wire, "Pivotal Manufacturing Partners and Meadow Partners Acquire Advanced Manufacturing Facility in Greater Phoenix," August 10, 2026 (53 Ten, 157,523 square feet, Glendale, Corning affiliate tenant)
Frequently Asked Questions
- Does reshoring manufacturing to the US make sense on cost?
- Mostly no. On comparative advantage the US can buy chips, drugs, and magnets more cheaply from Asia, and Kearney's Reshoring Index stayed negative in 2025. The case for reshoring is supply security, not cost.
- Then why reshore if it is more expensive?
- Because a few supply chains are concentrated in China, which has used export controls as leverage. China holds about 90 percent of rare earth refining and about 94 percent of 6-APA for amoxicillin. A cutoff would cost more than building at home.
- Is the US factory building boom still growing?
- No. It peaked near $249.8 billion in 2024 and fell 22.1 percent year to date through June 2026 as CHIPS-era subsidies rolled off. Data center construction passed semiconductor fab construction in spring 2026.
- How does reshoring affect commercial real estate investors?
- It shows up as concentrated demand for advanced manufacturing and supplier buildings near a few anchors, not broad industrial demand. CBRE flags Phoenix, anchored by TSMC and Intel, as a market where available space may tighten in 2026.
- Is Phoenix benefiting from reshoring?
- Yes. TSMC's Arizona commitment totals about $265 billion and Intel's Fab 52 anchors supplier in-migration. In 2026 investors bought 53 Ten, a 157,523 square foot advanced manufacturing building in Glendale leased to a Corning affiliate serving semiconductor, aerospace, and defense.