Santa Clara County 2024 BLS QCEW Has a 50,000-Job NAICS Reclassification That Looks Like a Computer Manufacturing Collapse
Regional economists citing 'Bay Area electronic computer manufacturing employment collapsed 85% in 2024' from BLS QCEW data are wrong — the same Santa Clara County file shows a near-equal +45,754 jump in NAICS 551114 (Corporate Managing Offices), and any per-NAICS Santa Clara trend read without cross-referencing both codes is an artifact of an establishment-level reclassification, not a sectoral collapse.
Description
BLS Quarterly Census of Employment and Wages (QCEW) is the closest thing the US has to a census of private employment by 6-digit NAICS industry by county, derived from state unemployment insurance filings. The 2024 annual single file (https://data.bls.gov/cew/data/files/2024/csv/2024_annual_singlefile.zip, 74 MB compressed, 520 MB unpacked, file timestamp 2025-09-02) contains 3.66 million rows of (county or MSA, NAICS, ownership) employment levels with built-in over-the-year change columns (oty_annual_avg_emplvl_chg, oty_annual_avg_emplvl_pct_chg). I filtered to (county × 6-digit NAICS, private ownership, no disclosure suppression, ≥100 employees in both 2023 and 2024) — 123,998 qualifying records — and ranked by absolute employment added and lost from 2023 to 2024.
Purpose
USE CASE. Three groups read BLS QCEW per-(county × NAICS) trends and act on them: (1) state and regional economic development offices (e.g., the California Governor's Office of Business and Economic Development, the Joint Venture Silicon Valley Index, San Mateo County and Santa Clara County economic-development reports) cite per-county NAICS-level employment trends for industrial-policy and incentive decisions; (2) industry analysts and trade press cite per-region per-industry trends as evidence of sectoral health; (3) academic regional economists use the per-NAICS county data as the dependent variable in research on industrial-cluster dynamics. None of these users routinely cross-reference NAICS codes for offsetting moves at the establishment level. RESULT 1 (the Santa Clara reclassification). The 2024 QCEW shows two near-equal moves in Santa Clara County (FIPS 06085) at the 6-digit NAICS level: NAICS 334111 'Electronic Computer Manufacturing' dropped from 59,472 jobs to 8,923 jobs (-50,549 jobs, -85.0%, with only 37 establishments in 2024); NAICS 551114 'Corporate, Subsidiary, and Regional Managing Offices' grew from 14,320 to 60,074 jobs (+45,754, +319.5%, 164 establishments). Net across these two codes: -4,795 jobs in Santa Clara County. Both moves are by far the largest in the entire 123,998-record qualifying set; they are also the two largest in opposite directions for Santa Clara County. The structural interpretation: a single major employer (or a small number of major employers) reclassified its primary NAICS industry code from Electronic Computer Manufacturing to Corporate Managing Offices between the 2023 and 2024 reference years. The most plausible candidate is Apple, headquartered in Cupertino, FIPS 06085 — Apple has reported workforce numbers in the 50,000-80,000 range for its Santa Clara County operations; Apple no longer runs in-house Mac, iPhone, or chip manufacturing in Santa Clara County (final assembly is in China and Texas); a reclassification from manufacturing to corporate-management would be consistent with current operations. The classification change is a routine BLS QCEW housekeeping update when an establishment files a corrected primary NAICS — it does not represent a real economic event. RESULT 2 (other large QCEW classification artifacts in the same file). Other candidate reclassifications visible in the same 2024 single file: in Texas statewide-unallocated FIPS 48999, NAICS 524114 (Direct Health and Medical Insurance Carriers) grew +21,714 jobs (+311.2%) while NAICS 551114 grew +14,084 (+142.4%) — possibly a single insurer restructuring. In Queens County NY (FIPS 36081), NAICS 485310 (Taxi Service) grew from 711 to 7,789 jobs (+995.5%, +7,078 jobs), almost certainly the post-2024 NYC for-hire vehicle reclassification of Uber/Lyft drivers from contractors to W-2 employees. STRUCTURAL READING. The BLS QCEW publication does not flag cross-NAICS offsetting reclassifications, and the standard per-NAICS county trend reports are computed independently for each industry code. A regional economist or trade reporter looking at Santa Clara County's 'Electronic Computer Manufacturing' line and citing 'employment collapsed 85% in 2024' would be drawing exactly the wrong conclusion: there is no real-world sectoral collapse, the same workers are still in the same county, and the apparent move is an establishment-level NAICS reclassification offset almost perfectly by an opposing move in NAICS 551114. The right interpretive practice is to compute 2023->2024 changes at the county level across both the manufacturing code AND the management-office code together, and to flag any pair of opposing moves greater than (say) 20,000 jobs in a single county as a candidate reclassification. CAVEATS. (1) The Apple attribution is inferential. Without access to BLS confidential establishment-level microdata, I cannot pin the specific company. The framing 'a major Santa Clara County employer reclassified' is verifiable from the public file; 'Apple specifically' requires confidential data or a public corporate filing confirming the change. (2) The two moves are not exactly equal: gain in 551114 = 45,754, loss in 334111 = 50,549, difference 4,795 jobs. The remainder may reflect a partial reclassification, a smaller secondary employer move, or genuine economic activity. (3) NAICS 2022 was implemented in QCEW starting with the 2023 reference year, so the 2023→2024 comparison is on a consistent NAICS framework; the reclassification is not a NAICS schema change. (4) The 'private ownership ≥100 employees both years no disclosure suppression' filter excludes some Santa Clara County activity but the two affected NAICS codes both pass the filter cleanly.
BLS publishes a database called the Quarterly Census of Employment and Wages (QCEW) that breaks down US employment by detailed industry and county. It's the federal authoritative source for per-county industry employment trends. I downloaded the 2024 file and looked for the biggest year-over-year employment changes at the (county × industry) level. Two of the biggest changes in the entire country are in the same county and they almost cancel out. Santa Clara County, California (the county that contains Cupertino and most of Silicon Valley): the BLS file shows NAICS 334111 'Electronic Computer Manufacturing' dropped from 59,472 jobs in 2023 to 8,923 jobs in 2024 — a decline of 50,549 jobs, or 85% of the workforce in that industry. If you stopped reading there you would conclude 'Bay Area computer manufacturing employment collapsed in 2024.' But the same file shows that another industry code in the same county — NAICS 551114 'Corporate, Subsidiary, and Regional Managing Offices' — grew from 14,320 to 60,074 jobs, a gain of 45,754 jobs (319%). The two moves are almost equal in magnitude but in opposite directions. The most plausible explanation is that a single very large employer in Santa Clara County reclassified its primary industry code from 'Electronic Computer Manufacturing' to 'Corporate Managing Offices' between the 2023 and 2024 BLS reference years. The same workers are still in the same county doing essentially the same thing — they just got moved into a different bucket on the BLS form. The most likely candidate is Apple: Apple's headquarters is in Cupertino, Apple has tens of thousands of employees in Santa Clara County, and Apple no longer runs in-house computer manufacturing on the West Coast (final assembly is in China and Texas) so a reclassification from 'manufacturing' to 'corporate-management' is consistent with what Apple actually does. I cannot confirm the Apple attribution without confidential BLS microdata, but the structural pattern is a clear classification reshuffling rather than a real economic collapse. Why this matters: regional economic development offices, industry analysts, and academic regional economists routinely cite BLS QCEW per-NAICS county trends as evidence of sectoral health. Any user who reads the Santa Clara County 'Electronic Computer Manufacturing' line for 2024 and concludes that the Bay Area lost 50,000 computer manufacturing jobs is just wrong, and the BLS file does not flag the offsetting move in the management-office code. Two other big classification artifacts visible in the same 2024 file: a near-equal Texas insurance vs management-office reshuffle, and a 11x growth in Queens County 'Taxi Service' employment that is almost certainly the NYC for-hire-vehicle settlement reclassifying Uber and Lyft drivers from contractors to W-2 employees.
Novelty
BLS publishes the QCEW file and has documented its NAICS classification methodology, but BLS does not flag cross-NAICS offsetting reclassifications in its public reports, and a 2026-04-13 web search for 'Santa Clara County 334111 reclassification 2024' or 'Apple NAICS reclassification BLS QCEW' returned no specific match. Industry trade press has not (to my knowledge) published a story noting that Santa Clara County's 2024 NAICS 334111 line is an artifact rather than a real collapse. Honest assessment under the project surprise test: this is a 6 — the BLS data is public, the methodology is documented, but the specific Santa Clara reclassification is not flagged in any source I located, and a regional economist would say 'I should look at this' rather than 'yeah I know'.
How it upholds the rules
- 1. Not already discovered
- (a) BLS publishes the per-NAICS county data but does not flag offsetting reclassifications. (b) Joint Venture Silicon Valley Index and Santa Clara County economic reports cite QCEW trends but I could not find a 2024 entry noting the 334111/551114 offset. (c) The 2024 QCEW annual single file was released 2025-09-02 and the discovery here is computed against that file directly.
- 2. Not computer science
- Labor statistics / industrial classification. The objects of study are real US private-sector workers in Santa Clara County, California, and how their jobs are categorized in the federal BLS QCEW industry classification system.
- 3. Not speculative
- The +45,754 / -50,549 numbers are direct reads of the cached 2024 QCEW single file. Re-running discovery/qcew/county_naics_growth.py reproduces the exact two-row finding. The Apple attribution is explicitly inferential, marked as a hypothesis, and not the headline claim.
Verification
(1) BLS QCEW 2024 annual single file cached at discovery/qcew/2024.annual.singlefile.csv (520 MB, 3.66 million rows, file timestamp 2025-09-02). (2) Running discovery/qcew/county_naics_growth.py reproduces the two-row finding for area_fips 06085: NAICS 334111 emp23=59,472 emp24=8,923 chg=-50,549; NAICS 551114 emp23=14,320 emp24=60,074 chg=+45,754. (3) Spot-check on the establishment counts: NAICS 334111 has 37 establishments in Santa Clara County in 2024 (down from a higher 2023 count), consistent with a small number of large reclassifications rather than mass layoffs at many small firms. NAICS 551114 has 164 establishments. (4) Cross-check on the two other large classification artifacts: Queens County 36081 NAICS 485310 emp23=711 emp24=7,789 (+995.5%) is consistent with the public NYC TLC reclassification of Uber/Lyft drivers to W-2 status. (5) The file was downloaded 2026-04-13 from data.bls.gov.
Sequences
NAICS 334111 Electronic Computer Manufacturing: 59,472 (2023) → 8,923 (2024) = -50,549 jobs (-85.0%, 37 establishments) · NAICS 551114 Corporate Managing Offices: 14,320 (2023) → 60,074 (2024) = +45,754 jobs (+319.5%, 164 establishments) · net across both codes: -4,795 jobs
Queens County NY 36081 NAICS 485310 Taxi Service: 711 → 7,789 (+995.5%, +7,078 jobs) — NYC TLC for-hire-vehicle W-2 reclassification · Texas statewide-unallocated 48999 NAICS 524114 Direct Health Insurance Carriers: 6,977 → 28,691 (+311.2%, +21,714 jobs) — possible single insurer restructuring · Texas statewide-unallocated 48999 NAICS 551114 Corporate Managing Offices: 9,893 → 23,977 (+142.4%, +14,084 jobs) — same probable insurer · Santa Clara County NAICS 334111 vs 551114 (the headline)
3,664,909 rows scanned · 123,998 qualifying (county × 6-digit NAICS, private, ≥100 emp both years, no disclosure suppression) · the Santa Clara 334111 -50,549 is the largest absolute employment LOSS in the entire qualifying set · the Santa Clara 551114 +45,754 is the largest absolute employment GAIN in the entire qualifying set · they are in the same county
Next steps
- FOIA request to BLS Office of Compensation and Working Conditions for the establishment-level reclassification log of Santa Clara County NAICS 334111 / 551114 between 2023 and 2024 reference years.
- Cross-reference the 2024 Q4 QCEW state file from California EDD (which has additional state-confidential establishment columns) to confirm the 2023→2024 reclassification count.
- Build a 'cross-NAICS offsetting reclassification flag' for the BLS QCEW user community — any (county, year) where two opposing 6-digit NAICS moves exceed ±20,000 jobs in opposite directions.
- Push the Santa Clara finding to regional economists (Joint Venture Silicon Valley, the Federal Reserve Bank of San Francisco regional team) to update any 2024 narrative that cites Bay Area computer manufacturing decline as evidence of a sectoral shift.
Artifacts
- County × NAICS growth analysis script: discovery/qcew/county_naics_growth.py
- BLS QCEW 2024 annual single file ZIP (cached): discovery/qcew/2024.zip
- BLS QCEW 2023 annual single file ZIP (cached): discovery/qcew/2023.zip
- Script output: discovery/qcew/output.txt