Producing ten major consumer-technology categories entirely in the United States by 2031 would require $185–$230 billion of capital spending, 555,000–668,000 additional workers and about 19.1–19.5 terawatt-hours of electricity each year, according to a Consumer Technology Association study prepared with Kearney. The scale makes a phased assembly strategy economically different from complete supply-chain localization.
The model covers smartphones, laptops, smartwatches, televisions, game consoles, monitors, smart speakers, robotic vacuums, wireless earbuds and headphones. It compares full U.S. production with a scenario in which final assembly moves domestically while most components continue to come from global suppliers.
The assembly alternative is much smaller
CTA estimates assembly-only reshoring would require $16–$19 billion of capital, 61,000–73,000 workers and 1.4–1.8 terawatt-hours of annual electricity. Using the midpoints of the published ranges, full production requires roughly 11.9 times the capital, 9.1 times the labor and 12.1 times the power.
That gap explains why policy language about “making technology in America” needs a precise definition. Moving final assembly can create domestic jobs and reduce some logistics exposure without immediately duplicating every upstream plant. Rebuilding processor, memory, display, battery and component ecosystems is a far larger industrial project.
The midpoint capital estimates imply about $342,000 per added worker for full production versus roughly $261,000 for assembly alone. Those rough ratios are not project budgets, because the categories and timing differ, but they show that the upstream scenario is more capital intensive as well as larger. Financing costs and depreciation would therefore become material inputs to product economics long before every factory reached mature utilization.
Consumers would bear part of the transition
CTA's model estimates that full domestic production would raise manufacturing costs by 152% for smartphones, 97% for smartwatches, 93% for laptops and 41% for televisions. Assuming companies pass through 25%–50% of the increase, the modeled retail-price effect is 38%–76% for smartphones, 23%–46% for laptops and 10%–21% for televisions.
Manufacturing-cost changes and retail-price changes are not interchangeable. Companies can absorb some expense through margins, product redesign or supplier negotiation, while consumers may delay replacements when prices rise. The estimates also use tariffs in force during September 2026, so policy changes would alter the result.
Demand elasticity is the missing bridge. A 38%–76% modeled smartphone price increase does not mean revenue rises by the same amount: buyers could keep devices longer, trade down or shift purchases to categories with more imported content. Producers could respond by reducing features, accepting lower margins or localizing only selected components. The eventual burden would be divided among consumers, manufacturers, suppliers and taxpayers rather than landing in one place.
Energy and labor are binding constraints
Full reshoring would require more than double the existing U.S. computer-and-electronics manufacturing workforce in the association's analysis. It would also compete for electricity with data centers and other industrial projects. Capital subsidies cannot instantly create skilled workers, grid connections or supplier expertise.
Grid timing may be as important as generation volume. New factories and data centers need transmission, substations and interconnection approvals at specific locations; national annual-energy totals do not guarantee that power is available where a plant is built. The same local constraint applies to housing and training for hundreds of thousands of workers. A subsidy can improve a project's financial return without removing those physical bottlenecks.
The White House's Genesis Mission supports AI-enabled science and advanced technologies through federal coordination and more than $5 billion of announced commitments. It may improve research and industrial capability, but it is not a funded plan to replicate the ten consumer-product supply chains measured by CTA.
CTA represents the technology industry, and the estimates are modeled rather than observed costs. Even with that caveat, the order-of-magnitude difference supports a practical conclusion: assembly-first localization and diversified trusted suppliers can reduce exposure sooner, while full domestic production is a multiyear capacity build with inflation, margin and execution consequences.