Charted: Manufacturing Led 2024 Idle Days at 7.7% Density — Utilities at 17.8% Logged Zero

Major work stoppages are rare enough that a handful of contracts can rewrite the national idle-day ledger — and common enough that desks still treat “high union density” as a proxy for “where the strikes are.” That proxy fails the 2024–2025 cross-section. The Bureau of Labor Statistics Work Stoppages Program counts stoppages involving 1,000 or more workers lasting at least one weekday shift [BLS WSP]. Overlay those idle days on CPS union membership and representation (coverage) rates, and the map is a scatter, not a staircase.

The interactive dashboard above pairs 2024 industry idle days (BLS Economics Daily industry table) with 2025 CPS coverage (Union Members Table 3). Toggle Scatter, Rank, Timeline, and Ownership; switch the metric between days idle and workers idled; filter goods, services, or infrastructure bands. The headline is deliberately two-sided: manufacturing logged 1.32 million cumulative days idle at 7.7% membership, while utilities — the highest major private membership rate at 17.8% — do not appear on the 2024 major-idle industry table at all [BLS WSP 2024] [BLS Union Members 2025].

Scoreboard: idle days versus coverage

Read the table as a mismatch ledger. Education led headcount; manufacturing led duration-weighted idle days; utilities led private density without a 2024 major-idle footprint. High coverage is neither necessary nor sufficient for a large idle-day print.

What BLS counts — and what it does not

A “day of idleness” is a scheduled weekday (Monday–Friday, excluding federal holidays) when a counted worker does not work because of a major stoppage. BLS multiplies workers by lost workdays in the reference month [BLS WSP]. The program does not separate strikes from lockouts, and it excludes secondary idleness at suppliers or customers. Only disputes that clear the 1,000-worker threshold enter the major series. Shop-floor walkouts below that bar — and many short private disputes — never hit this dashboard.

Union rates come from a different survey. The CPS Union Members release measures wage-and-salary workers who are members or represented by a union on their principal job. Representation is the broader “coverage” concept used on the scatter’s X-axis [BLS Union Members 2025]. For 2025, BLS warns that annual averages are 11-month figures excluding October (federal shutdown), so year-to-year CPS comparisons are not strictly clean. That caveat matters for trend chatter; it does not erase the cross-industry ranking that puts utilities atop private membership and finance near the floor.

The scatter: mid-coverage industries carried the idle-day mass

If idle days simply followed density, the ≥12% coverage cluster (education, construction, transportation, utilities) would dominate the Y-axis. It does not. Summing 2024 idle days into coverage buckets shows the 5–12% band — manufacturing, information, health care — holding more cumulative idle days than the high-coverage band. Education is the high-density exception that actually shows up: 782,200 idle days at 15.7% coverage. Transportation sits at 14.9% coverage with only 28,000 idle days. Utilities sit at 18.8% coverage with a disclosed zero on the major-idle table.

Information is the other instructive outlier. At 6.6% membership — near the private average — it still produced 620,800 idle days in 2024, a duration story more than a density story. Accommodation and food services deepen the point from the other side: 2.4% membership and still 236,200 idle days. Low coverage does not mean an industry is strike-proof on the major series; it means fewer workers are covered, not that every covered dispute is short.

Education leads people; manufacturing leads days

Workers idled and days idle answer different operational questions. In 2024, educational services involved 110,400 workers — the largest headcount on the industry table — while manufacturing’s 38,000 workers generated 1.32 million days idle [BLS WSP 2024]. Average duration (days idle ÷ workers) is a desk ratio, not a BLS product, but the arithmetic is blunt: manufacturing disputes that cleared the major threshold lasted much longer on a worker-weighted basis than the education cluster.

That split also explains why ownership totals can look paradoxical. In 2024, private employers accounted for most idle days (3.0 million) even though public-sector membership (32.2% in 2024, 32.9% in 2025) is more than five times the private rate (5.9%) [BLS WSP 2024] [BLS Union Members 2025]. Dense public bargaining units can idle many people for shorter calendars; a smaller set of private industrial and media stoppages can dominate the day-count. In 2025 the ownership balance flipped on headcount — state and local government idled 232,800 workers versus 74,000 in private industry — while cumulative days idle stayed closer (711,600 government vs 784,400 private) [BLS WSP 2025]. Density shapes who can strike; contract calendars and industry structure shape how many days the ledger records.

The 2023 spike is not the 2025 baseline

Annual idle days jumped to 16.7 million in 2023 — an order of magnitude above the mid-2010s quiet years — then fell to 3.36 million in 2024 and 1.99 million in 2025 [BLS WSP]. Stoppages beginning in the year stayed near 30–33 across 2023–2025; the volatility is duration and who is idled, not a simple explosion in the count of major disputes. The 2023 industry headcount table is dominated by information (171,500 workers), then health care and education, with manufacturing at 61,200 [BLS WSP 2023]. Overlay 2025 membership rates and again the ranking is not density-ordered: information and health care out-idled education on workers despite lower membership rates.

Treat 2023 as a spike year, not a new steady state. The dashboard’s timeline panel shows stoppage counts recovering from the 2020 trough while idle-day totals remain noisy. A single long multi-employer stoppage can overwhelm a year of shorter public disputes. That is why “are strikes happening where density is high?” needs a year label and a metric label before it gets a yes or no.

2025: service sectors own the worker share

For stoppages beginning in 2025, BLS reports 306,800 workers idled and 1.99 million cumulative days idle. Service-providing industries accounted for 98% of those workers: education and health services alone 196,500 (64%), public administration 82,300, other services 21,800, and manufacturing only 6,200 (2%) [BLS WSP 2025]. The worker-share chart is therefore a public-services story in 2025 — consistent with elevated public density — even as the private idle-day total still slightly exceeds the state/local day total.

That is the cleanest way to reconcile the two series. Coverage is highest in government and a few private infrastructure niches. Major idle days in recent years have clustered in education, health, information, and manufacturing — only one of which sits in the top private density tier. Asking whether strikes “happen where density is already high” gets a partial yes for education and public administration, and a clear no for utilities-versus-manufacturing and for accommodation-and-food’s low-density idle print.

Caveats and reading rules

• Threshold bias: The major series ignores disputes under 1,000 workers. Industry rankings can miss fragmented private organizing.
• Strike vs lockout: BLS combines both; idle days are not a pure “worker power” index.
• Secondary effects: Supplier and customer downtime is out of scope.
• CPS vs WSP: Membership/coverage and idle days come from different instruments and reference periods; the scatter is a joint desk construction.
• 2025 CPS: Eleven-month averages excluding October; use levels more carefully than one-year deltas.
• Utilities zero: Absence from the 2024 industry idle-day table is treated as zero major idle days for that published cutnot proof that no utility dispute existed below the major threshold.
• Occupation vs industry: Teachers drive education’s density; CPS industry rows mix public and private employers inside some service sectors.

Bottom line for the density question

Union coverage tells you where contracts and dues already sit. Major-work-stoppage idle days tell you where large, recorded disputes burned calendar time. In the latest clean industry cross-section, those two maps rhyme only in places — education above all — and diverge sharply where utilities’ peak private density meets a blank idle-day cell and manufacturing’s mid-single-digit density meets the largest idle-day total. Strikes are not a simple function of how unionized an industry already is; they are a function of bargaining calendars, unit size, duration, and which disputes clear BLS’s major gate.