Nevada, Florida, Maryland, Texas, and Arizona are most exposed if foreign-born construction labor were disrupted; nationwide, 26.3% of construction workers are foreign-born

Summary

Foreign-born workers made up 26.3% of the U.S. construction workforce in 2024, a record high, and about 3 million people nationally (NAHB/Eye on Housing, from Census ACS microdata). Dependence is heavily concentrated: California (42%), Florida (41%), Texas (39%), Nevada (38%), Maryland and New Jersey (37% each), and New York (37%) lead the country, while a cluster of low-immigration states -- New Hampshire, Montana, Alaska, West Virginia, Vermont -- sit under 5%. Within the trades that build houses specifically, dependence is even higher: 57% of drywall/ceiling-tile installers, 56% of plasterers/stucco masons, 53% of roofers and painters, and 43% of construction laborers are foreign-born. Combining these shares with each state's construction share of GDP (BEA, queried directly) to build a simple exposure index, Nevada, Florida, Maryland, Texas, and Arizona stand out as the states where a foreign-born labor-supply shock would put the largest fraction of total state economic output at risk -- because they combine a high foreign-born share of the construction trades with construction being an outsized part of their state economy. California and New York, despite very large numbers of affected workers, rank lower on this output-exposure measure because construction is a comparatively smaller slice of their much larger, more diversified economies.

1. National and state-level dependence on foreign-born labor

Source: NAHB analysis of 2024 American Community Survey (ACS) data, published by Eye on Housing (NAHB's research blog), April 2026 -- read directly via the primary article.

• National: 26.3% of the construction workforce is foreign-born (a record), rising to about one in three among the 28 specific construction trade occupations (which make up ~60% of the industry's labor force).
• Highest-dependence states: California 42%, Florida 41%, Texas 39%, Nevada 38%, Maryland 37%, New Jersey 37%, New York 37%.
• Next band (>25%, exact figure not published in the source): Connecticut, Massachusetts, Georgia, Virginia, Illinois, Arizona, North Carolina.
• Lowest-dependence states (<5%): New Hampshire, Montana, Alaska, West Virginia, Vermont.
• Scale: California and Texas each have more than half a million foreign-born construction workers; together with Florida and New York they account for 52% of the nation's roughly 3 million immigrant construction workers.
• Within trades essential to home building specifically, foreign-born concentration runs well above the industry-wide average: drywall/ceiling-tile installers 57%, plasterers/stucco masons 56%, roofers 53%, painters 53%, carpet/floor/tile installers 51%, construction laborers 43%, carpenters 35%.
• Note on definitions: ACS classifies as "foreign-born" both naturalized U.S. citizens and non-citizens (which in turn includes green-card holders, work-visa holders, and the unauthorized population) -- it does not record legal status, so this figure cannot be split into a documented/undocumented breakdown.

2. Where a labor-supply shock would hit construction output hardest

Workforce share alone does not tell you where the ECONOMY is most exposed -- a state can have a high foreign-born share of its construction workforce but a construction sector that is a small part of an otherwise huge, diversified economy (California, New York), or vice versa. To locate output exposure, this analysis queried BEA's SAGDP2 current-dollar state GDP-by-industry table (econ.state_gdp in this warehouse, table_name='SAGDP2', line_code 11 = Construction vs. line_code 1 = All industry total, year 2024) for construction's share of each state's total GDP, then multiplied that share by the foreign-born share of that state's construction workforce (from the NAHB/ACS figures above). The result approximates what fraction of a state's total economic output is construction output produced by foreign-born labor -- the output most directly exposed if that labor supply were disrupted.

Ranked highest to lowest exposure among the states with published foreign-born shares: Nevada (2.94), Florida (2.26), Maryland (2.18), Texas (1.93), Arizona (1.67), California (1.62), Georgia (1.35), New Jersey (1.30), Virginia and North Carolina (1.27 each) -- all above the U.S. average of 1.17. New York (1.07), Massachusetts (1.00), Illinois (0.93), and Connecticut (0.77) sit near or below the national figure despite high or elevated workforce shares, because construction is a smaller share of their overall state GDP (queried directly: NY 2.9% of GDP, IL 3.6%, vs. Nevada's 7.7%). The five lowest-dependence states (New Hampshire, Vermont, West Virginia, Alaska, Montana) also register the lowest exposure, at roughly a sixth to a third of the national figure.

Mechanically, this means: a broad disruption to foreign-born construction labor (e.g., a large-scale immigration enforcement action, a visa-program change, or a border/labor-supply shock) would be expected to hit Nevada, Florida, Maryland, Texas, and Arizona hardest in terms of the SHARE of their state economies affected, even though California and Texas would see the largest ABSOLUTE number of workers affected (over half a million each). Nevada's position at the top is driven less by an extreme workforce share (38%, fourth-highest) and more by construction being unusually large relative to its state economy (7.7% of state GDP in 2024, well above the ~4.5% national average, both from the same BEA query) -- consistent with Nevada's economy being heavily tied to new construction (Las Vegas-area residential and hospitality building).

3. What this exposure index does NOT capture

• It is a static output-share proxy, not a modeled shock. It does not estimate how much construction output would actually fall following a given labor-supply reduction (that requires an elasticity of output with respect to labor input, wage/price responses, and how quickly native-born or other-immigrant labor could substitute in -- none of which this index models).
• The mid-band states (CT, MA, GA, VA, IL, AZ, NC) use a >25% floor, not their true value, because NAHB's public write-up does not publish exact percentages for that band. Their true exposure index is understated to some degree if their actual share is well above 25% (e.g., closer to 35%, as it plausibly is for Georgia and Arizona given their construction booms).
• The state-by-state foreign-born-by-industry breakdown itself came entirely from the NAHB/Eye on Housing article, not from a query run in this session. This warehouse's Census ACS tables carry nativity and industry as separate published aggregates (population by nativity, and employment by industry), not a joint nativity-by-industry-by-state cross-tabulation, so the state figures above could not be independently recomputed here from raw ACS data -- they rely entirely on NAHB's own tabulation (built from ACS PUMS microdata, which does support that cross-tab). This is a genuine sourcing gap in what could be queried directly, not a shortcut taken to avoid the work. Only the construction-share-of-state-GDP half of the exposure index was computed from a live query in this session.
• Homebuilding specifically vs. "construction" broadly: the NAICS 23 construction sector spans residential, nonresidential, and heavy/civil construction plus specialty trade contractors. NAHB's trade-level breakdown (drywall, roofing, painting, etc.) is the closer proxy for pure homebuilding exposure, and those trade-specific foreign-born shares run higher (51-57%) than the all-construction state averages used in the exposure index -- meaning the index above, if anything, UNDERSTATES exposure for the trades most specific to home construction.

What This Report Does Not Answer

• Coverage should be near-exhaustive across states, not a few examples: NAHB/Eye on Housing's public report gives exact foreign-born shares for ~19 states (the highest and lowest bands) and lumps mid-tier states into a '>25%' band; it does not publish an exact figure for all 50 states, and this warehouse could not independently recompute the missing states because no queried table here crosses nativity by industry by state. Disclosed as a genuine data gap, not glossed over.

Every query behind this report

9 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT DISTINCT table_name, line_code, description FROM econ.state_gdp WHERE description ILIKE '%construction%' AND geo_name='United States' ORDER BY table_name, line_code

SELECT DISTINCT table_name, line_code, description FROM econ.state_gdp WHERE description ILIKE '%all industry total%' AND geo_name='United States' ORDER BY table_name, line_code

SELECT DISTINCT table_name, unit FROM econ.state_gdp WHERE table_name IN ('SAGDP8','SAGDP9') LIMIT 10

SELECT DISTINCT table_name FROM econ.state_gdp

SELECT DISTINCT table_name, line_code, description, unit FROM econ.state_gdp WHERE table_name='SAGDP2' AND (description ILIKE '%construction%' OR description ILIKE '%all industry total%') AND geo_name='United States'

SELECT DISTINCT line_code, description, unit FROM econ.state_gdp WHERE table_name='SAGDP2' AND geo_name='United States' ORDER BY line_code FETCH FIRST 30 ROWS ONLY

SELECT DISTINCT geo_name FROM econ.state_gdp WHERE table_name='SAGDP2' FETCH FIRST 10 ROWS ONLY

SELECT DISTINCT line_code, description, unit FROM econ.state_gdp WHERE table_name='SAGDP2' AND geo_name='Texas' ORDER BY CAST(line_code AS INTEGER) FETCH FIRST 40 ROWS ONLY

SELECT geo_name, year,
  MAX(CASE WHEN line_code='11' THEN data_value END) AS construction_gdp,
  MAX(CASE WHEN line_code='1' THEN data_value END) AS total_gdp
FROM econ.state_gdp
WHERE table_name='SAGDP2' AND year=2024
GROUP BY geo_name, year
ORDER BY geo_name