DC Market Panel how tight is the labor where you're building?
State resolution averages Loudoun with Bristol. This is construction wages and headcount where the shovels are — tight core counties for 20 real data-center markets, measured against the national rate. Craft labor is the constraint nobody prices until it bites: a market adding construction workers twice as fast as the country is a market where your subcontractor coverage is thinning.
Export data
QCEW quarter 2026-01-01 vs 2025-01-01 · roster curated 2026-10-04. QCEW publishes ~7 months after quarter end — these are the freshest county wages that exist, not a current reading.
Where construction labor is tightest
Each market's wage growth above the national rate, plus half its headcount growth above the national rate, in percentage points. Ten or more is hot; three or more is warm.
- Hot
- Warm
- Neutral
- Slack
Not ranked: Hillsboro OR — BLS withholds these counties' construction figures, or there is no year-earlier quarter to compare.
Every market, county by county sort any column; open a row for the per-county receipts
| Tightness | |||||||
|---|---|---|---|---|---|---|---|
MISO · Entergy Louisiana | Hot | $2,559 | +181.5%+177.1pp vs US | 2,052 | +968.8%+967.8pp vs US | 1,440 MW under constr. 1 tracked site | — BLS-suppressed (Richland Parish) |
ERCOT · AEP Texas | Hot | $1,522 | +21.5%+17.1pp vs US | 4,752 | +37.7%+36.7pp vs US | 900 MW under constr. 2 tracked sites · 300 MW operating | +19.4%+12.7pp vs US 497 workers · $1,536/wk |
WECC · NV Energy | Hot | $1,746 | +17.4%+13.0pp vs US | 24,044 | +7.9%+6.9pp vs US | — 0 tracked sites | +24.6%+17.9pp vs US † 2,096 workers · $2,342/wk |
WECC · Black Hills Energy | Warm | $1,503 | +11.7%+7.3pp vs US | 3,962 | +2.5%+1.5pp vs US | — 2 tracked sites · MW not disclosed at 2 sites | +21%+14.3pp vs US 765 workers · $1,691/wk |
MISO · AEP Indiana Michigan | Hot | $1,643 | +11.5%+7.1pp vs US | 6,550 | +12.9%+11.9pp vs US | 0 MW under constr. 1 tracked site · 1,725 MW operating | +12.9%+6.2pp vs US 1,569 workers · $2,085/wk |
MISO · MidAmerican Energy | Hot | $1,380 | +8.8%+4.4pp vs US | 2,573 | +26.2%+25.2pp vs US | 0 MW under constr. 1 tracked site · 500 MW operating | +9.3%+2.6pp vs US 546 workers · $1,611/wk |
ERCOT · CPS Energy | Warm | $1,637 | +7.8%+3.4pp vs US | 53,268 | +6.7%+5.7pp vs US | — 0 tracked sites | +27.7%+21.0pp vs US 10,461 workers · $1,768/wk |
PJM · ComEd | Warm | $2,078 | +7.5%+3.1pp vs US | 101,266 | +1.8%+0.8pp vs US | — 1 tracked site · MW not disclosed at 1 site | +7.8%+1.1pp vs US 12,924 workers · $2,293/wk |
PJM · AEP Ohio | Warm | $1,928 | +7.3%+2.9pp vs US | 45,657 | +6.7%+5.7pp vs US | 0 MW under constr. 3 tracked sites · 1,221 MW operating | +19.8%+13.1pp vs US 8,391 workers · $2,270/wk |
PJM · Dominion Energy Virginia | Hot | $1,852 | +7.2%+2.8pp vs US | 44,761 | +15.4%+14.4pp vs US | — 1 tracked site · MW not disclosed at 1 site | +6.9%+0.2pp vs US 14,731 workers · $2,336/wk |
MISO · MidAmerican Energy | Warm | $1,722 | +5.8%+1.4pp vs US | 22,665 | +9.2%+8.2pp vs US | — 0 tracked sites | -2%−8.7pp vs US † 3,580 workers · $1,912/wk |
ERCOT · Oncor | Neutral | $1,845 | +5.5%+1.1pp vs US | 174,867 | +1.3%+0.3pp vs US | — 2 tracked sites · 250 MW operating · MW not disclosed at 1 site | +11.8%+5.1pp vs US 22,593 workers · $1,855/wk |
SERC · Georgia Power | Neutral | $2,371 | +5.5%+1.1pp vs US | 26,414 | -1.8%−2.8pp vs US | 200 MW under constr. 1 tracked site | +11.7%+5.0pp vs US 3,517 workers · $1,962/wk |
WECC · APS / SRP | Neutral | $1,742 | +5.1%+0.7pp vs US | 177,067 | -1.1%−2.1pp vs US | 0 MW under constr. 1 tracked site · 200 MW operating | +6.6%−0.1pp vs US 22,269 workers · $1,849/wk |
WECC · Rocky Mountain Power | Neutral | $1,567 | +4.8%+0.4pp vs US | 84,862 | -0.8%−1.8pp vs US | — 0 tracked sites | +1.1%−5.6pp vs US 6,435 workers · $1,927/wk |
CAISO · PG&E / Silicon Valley Power | Neutral | $2,295 | +4%−0.4pp vs US | 50,892 | +3.6%+2.6pp vs US | — 0 tracked sites | +1.3%−5.4pp vs US 7,535 workers · $3,150/wk |
WECC · Grant County PUD | Slack | $1,521 | +3%−1.4pp vs US | 1,572 | -3.5%−4.5pp vs US | — 2 tracked sites · MW not disclosed at 1 site · 18 MW planned | +1.4%−5.3pp vs US 196 workers · $1,968/wk |
TVA · MLGW | Slack | $1,561 | +0.5%−3.9pp vs US | 18,272 | +1.3%+0.3pp vs US | 0 MW under constr. 2 tracked sites · 1,400 MW operating | -0.9%−7.6pp vs US 2,952 workers · $1,618/wk |
MISO · We Energies | Slack | $1,543 | -8.5%−12.9pp vs US | 3,670 | -4.2%−5.2pp vs US | 0 MW under constr. 1 tracked site · 400 MW operating | -0.5%−7.2pp vs US 522 workers · $1,690/wk |
| Hillsboro OR | not available — BLS disclosure suppression (Washington County): Washington County — BLS disclosure-suppressed for private NAICS 23; retained so the suppression is visible rather than silently backfilled from Multnomah, which is a different labor market | +5.9%−0.8pp vs US 1,743 workers · $2,165/wk | |||||
How it's built
Wages and headcount
Wage is employment-weighted across each market's counties, and year-over-year uses a like-for-like county set: a county disclosure-suppressed in either quarter is excluded from both sides, so composition change can't contaminate the rate. Markets are tight core counties — where data centers actually are, not the metro area; per-county receipts expand on every row so the aggregation is checkable. Wage $/wk, Wage YoY and Headcount YoY stay on that like-for-like basis; Constr. workers is the market's full current-quarter headcount (emp_cur_total), independent of whether a county cleared last year's disclosure bar — expand a row for the reconciling current-quarter total and any counties the like-for-like receipts exclude (marked † when partial).
Tightness
Tightness buckets a composite score — the wage spread in percentage points plus half the employment spread — at ≥10 Hot, ≥3 Warm, and above −3 Neutral; −3 or below is Slack. The chart draws scores past 40 to the edge and prints the real value.
Capacity and attributes
MW under constr. counts only capacity tagged under construction at hand-curated sites; operating capacity is shown separately, beside the tracked-site count, because an energized campus is a completed draw on the labor pool, not a live one. Competition MW is drawn from the /capacity tracker: 27 public companies plus 2 private builders (xAI, OpenAI / Stargate), hand-curated from filings. Private operators (CyrusOne, Vantage, Aligned, STACK, QTS, EdgeConneX) and hyperscaler leased space inside their shells are not tracked, and sites with undisclosed locations carry no market. MW under construction counts only sites tagged st="c" -- operating, planned and secured MW are published separately, never folded in -- so treat it as a floor on what is actually being built, never a census. Utility and ISO are hand-curated attributes of the market, not derived.
Electrical contractors
Electrical contractors (nonres., NAICS 238212) is private-sector QCEW for nonresidential electrical contractors — NAICS 2022 split the old 238210 into 238211 (residential) and 238212 (nonresidential), and only the nonresidential trade, the one that wires commercial and data-center work, is shown. It is aggregated exactly like the construction columns: wage weighted by each county's quarterly-average employment, year-over-year on a like-for-like county set, workers as the current-quarter third-month level. The smaller figure beside the YoY is its spread against the national 238212 wage rate (+6.7% YoY, quarter 2026-01-01). A six-digit industry is often too small to publish in a single county: disclosure-suppressed counties are skipped, never zero-filled, a market missing any county for 238212 is marked †, and a market with none reads as suppressed. Expand a row for the per-county receipts.