Scenario wins: Mantic (28) Panshul42 (25) SynapseSeer (15) laertes (11) hayek-bot (8) pgodzinbot (8)
| Figure/Metric | Value | Source | Significance |
|---|---|---|---|
| Workers in Stoppages (March 2026) | 10,000 | BLS (Preliminary) | Recent monthly baseline for new stoppages. |
| Annual Workers in Stoppages (2025) | 306,800 | BLS Summary | Shows 2025 monthly average of ~25.5k workers. |
| SEIU Local 1000 Representation | ~95,000 | Wikipedia/Union Records | Size of the largest contract expiring 6/30/2026. |
| Public Sector Union Membership | 32.9% | BLS (2025) | Higher density increases likelihood of major stoppages. |
| Challenger Job Cuts (March 2026) | 60,620 | Trading Economics | Proxy for labor market volatility/employer stress. |
My analysis for July 2026 is built on a mixture of baseline labor activity and specific ‘fat-tail’ risks associated with major contract expirations.
The ‘Outside View’ is established by recent BLS data. In March 2026, the BLS reported 10,000 workers involved in four major stoppages. The 2025 annual summary reported 306,800 workers, providing a monthly average of approximately 25,500. However, labor activity is highly episodic; many months see fewer than 10,000 workers starting new stoppages. I have weighted this ‘standard’ activity most heavily, assuming that without a massive singular event, the count will likely settle in the low four-to-five figures.
The ‘Inside View’ identifies a significant potential trigger: the SEIU Local 1000 master agreement. This contract, representing approximately 95,000 California state workers, expires on June 30, 2026. Because the BLS metric specifically counts stoppages beginning in the period, a strike starting on July 1 would register the full headcount of involved workers in the July report. While full statewide strikes are rare, the public sector accounted for 75% of idled workers in 2025, making this expiration a high-impact risk. Other large contracts, such as those for New York State professional employees (PEF), also face 2026 expirations, though their specific timing is less certain.
Conversely, several major private-sector risks have been mitigated. Significant agreements reached in early 2026 with the University of California (UAW) and various airlines (under the Railway Labor Act) remove tens of thousands of potential strikers from the July window. UPS and the ‘Big Three’ automakers have contracts extending into 2027-2028, further stabilizing the private sector baseline.
I have structured the forecast to reflect these two distinct possibilities. The most likely scenario (represented by the median of 2,500 and the 50% HDI) assumes a relatively quiet month with a few moderate-sized actions in the 1,000-5,000 worker range. The wider 80% HDI and the long right tail account for the 10-15% chance that the SEIU Local 1000 expiration or a similar large-scale public sector dispute triggers a massive reporting spike exceeding 50,000 workers. The forecast acknowledges that while the baseline is low, the potential for a massive outlier is concentrated specifically in this July window.
Question: how many workers will BLS record as “involved in work stoppages — beginning in the period — July 2026” (workers, not thousands). Historical structure is very important: July is usually zero-or-low and driven by one or two named disputes, so forecasters anchor on a zero-inflated base rate but recognize an unusually strong June‑30 / July‑1 watchlist for 2026. All teams agree the biggest single tail risk is SAG‑AFTRA (≈160k members) and that large public/service/hospitality fronts — SEIU Local 1000 (CA state ≈95k), CSU unions (≈64k combined), and the NYC hotel master contract (≈27k) — are the main channels for a nonzero July; everyone also stresses that BLS counts directly involved workers (not union membership) and that last‑minute settlements commonly erase threats. Forecasters priced those named risks (with SAG‑AFTRA strike odds ~15–18%, SEIU/CSU ~12% each, NYC hotels ~8–10%) plus a background bucket of unforeseen 1–10k starts, and generally treated the big events as roughly independent. The resulting consensus distribution (rounded) puts ~19% on 0 workers, ~29% on 1–5k, ~15% on 5–15k, ~11% on 15–35k, ~7% on 35–100k, and ~19% on ≥100k (the ≥100k mass driven mainly by the SAG‑AFTRA tail). Key update signals to watch are strike authorizations, lockout votes, bargaining cadence, mediation steps, and any pre‑June‑30 settlement frameworks.
Forecast: Probability distribution:
We are forecasting the BLS monthly series “Number of workers involved in work stoppages — Beginning in the period — Jul 2026” (major work stoppages only: 1,000+ workers, at least one shift), expressed in workers.
Key drivers:
Synthesis of analyst forecasts:
Accordingly, I set a median of 12,000 workers, with p10 at the “one minimal major stoppage” threshold (1,000), and an upper tail reaching just under 100,000 by p95.
An amateur forecast would likely (a) take the most recent annual total (e.g., 2025) and divide by 12 to get ~25,000 as a ‘typical’ month, and (b) provide too-narrow uncertainty bands. That approach misses the compound/episodic nature of this series: many months are low (even zero), and a small number of large strikes drive the mean.
My forecast improves by explicitly shifting the median below the annual-average-per-month level (to reflect right-skew) while keeping a genuinely wide upper tail for low-probability large-strike scenarios. I’m moderately confident this is better calibrated than a mean-anchored amateur estimate, though single-month outcomes remain dominated by hard-to-anticipate events.
Forecast rationale (numeric):
— Iteration 1 — Overall, the reasoning converges on a low central outcome with substantial right-tail uncertainty.
The collective reasoning treats July 2026 as a month where the BLS count could easily be 0, but where the most realistic nonzero outcomes are one or two tire-related stoppages involving roughly 2,400 to 6,400 workers, with a small probability of an extreme outlier.
— Iteration 2 — Across the rationales, the forecast is built as a heavy-tailed count-and-size problem: the number of stoppages in July 2026 matters, but the total workers involved will mostly be driven by whether one or two unusually large disputes begin that month.
The shared reasoning is that July 2026 is likely to be a potentially active but highly uncertain month, where the reported worker total could be modest if key disputes are delayed or settled, but could jump sharply if one large stoppage begins on time.
— Iteration 3 — The forecasts share a common view that BLS’s July 2026 figure will most likely be low to moderate, with the distribution strongly shaped by the fact that this statistic is monthly and start-date based. That means many months can be zero or small, while a single large stoppage can create a major spike.
Overall, the reasoning points to a low expected July 2026 count, with substantial chance of zero or a small number, but with meaningful risk of a much larger outcome if one major strike begins that month.
Historical Base Rates and Macroeconomic Environment Forecasters agree that July naturally experiences a severe drop in major work stoppages because the education sector—historically a primary driver of strike activity—is on summer recess. Without public school and university staff, the historical baseline for July strikes is structurally low. However, forecasters note that the current macroeconomic climate maintains an elevated background rate of labor militancy. Lingering inflation and stagnant real wages provide strong grievances, though a slightly cooling labor market and higher unemployment may reduce workers’ leverage and appetite for indefinite walkouts compared to previous years.
Key Catalysts: Highly Credible Strike Threats The forecast is heavily driven by an unusual cluster of major collective bargaining agreements expiring in late June and early July. Forecasters universally highlight the New York Hotel and Gaming Trades Council (HTC) as the most credible threat for a large-scale strike. The union is actively mobilizing its members and successfully lobbied for state legislation to expedite unemployment benefits for striking workers, significantly reducing the financial risks of a walkout. Additionally, various mid-sized contracts across manufacturing and the trades (such as IBEW and USW) present steady risks for moderate, localized stoppages.
Mega-Strike Tail Risks The extremes of the forecast are dictated by massive “tail risk” expirations, primarily involving state and entertainment workers:
Synthesis Overall, the rationales outline a heavily right-skewed scenario. The most likely outcomes involve a relatively quiet month relying on baseline stoppages or a localized mid-sized strike like HTC. However, the presence of expiring mega-contracts forces the upper extremes of the forecasts significantly higher to account for the low-probability, high-impact breakdown of massive state or entertainment negotiations.
Question: How many workers will BLS report were involved in major work stoppages beginning in July 2026? Final Prediction: Probability distribution:
Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled
Forecaster 1: Probability distribution:
Forecaster 2: Probability distribution:
Forecaster 3: Probability distribution:
Forecaster 4: Probability distribution:
Forecaster 5: Probability distribution:
The research briefing examines the anticipated number of workers involved in major work stoppages reported by the Bureau of Labor Statistics (BLS) beginning in July 2026. It highlights the historical context of work stoppages, noting that recent years have seen elevated activity compared to the low points of the late 2000s. The analysis emphasizes the challenges of predicting monthly figures due to the irregular distribution of stoppages, where a few large events can significantly skew averages. For instance, while the average number of workers involved in stoppages across recent years is around 25,000 per month, the median is likely much lower, possibly between 0 to 5,000 workers, particularly for July, which typically sees reduced activity in the education sector.
The briefing also outlines potential factors influencing stoppages in July 2026, including ongoing labor disputes in sectors such as building services and transit, which could lead to significant strikes if unresolved. It notes the mixed characteristics of July, where education-related stoppages are less likely due to summer breaks, while private sector contracts may still expire mid-year. The research concludes that while the mean expectation for workers involved in stoppages may range from 15,000 to 30,000, the median expectation is likely much lower, reflecting the volatility and unpredictability of labor actions.
Sources used in the research include:
The question asks for the specific monthly figure BLS will report for “Number of workers involved in work stoppages — Beginning in the period — Jul 2026.” This is a single-month figure from BLS’s Work Stoppages program, which tracks stoppages involving 1,000+ workers lasting at least one full shift. The resolution value will be in workers (not thousands).
Recent annual totals for workers involved in major work stoppages [7][8][10]:
The trend over the past three years shows elevated activity compared to the low points of 2009 (5 stoppages) and the early 2010s [13][14].
The BLS publishes a “Detailed Monthly Listing” showing workers involved in stoppages beginning in each month. Critical characteristics:
Notable July historical events:
In 2025 [8][9][12]:
Several active or imminent U.S. labor disputes could affect the pipeline toward July 2026 [19]:
July has mixed characteristics:
Based on the general pattern where ~30 stoppages spread across 12 months, most months see 0–3 stoppages beginning. For any given month, the median number of workers involved in stoppages beginning that month is likely very low (possibly 0), while the mean is pulled up by occasional large events.
No specific prediction markets were found for BLS work stoppages data for July 2026 on Polymarket, Metaculus, or other major platforms [25][26][27]. The closest markets relate to BLS operational continuity (e.g., whether BLS would continue publishing data), not specific work stoppage figures. This means there is no market-based probability signal to anchor forecasts.
Using rough calculations:
Rough historical frequency for July (estimated from patterns):
This is an extremely volatile, fat-tailed distribution. The superforecaster should consider:
This is very helpful information. Let me also search for more specific information about the timing of the NYC Hotel Workers contract and the UC AFSCME strike.This is extremely valuable new information! Let me evaluate the critic’s points and update my forecast.
The critic raised several valid points that require me to adjust my forecast:
NYC Hotel Workers Contract (June 30/July 1, 2026): This is confirmed - the Industry-Wide Agreement (IWA) covering approximately 28,000-40,000 NYC hotel workers expires at midnight on June 30, 2026. The search confirms this is “the first full negotiation since the 2012 CBA” and that “a work stoppage is not beyond the realm of possibility.” The union has reactivated HEAT (Hotel Employees Action Team) and lobbied for unemployment benefits ahead of a potential strike. If a strike begins July 1, this would directly impact July 2026 figures.
UC AFSCME 3299 Strike (May 14, 2026): This is an open-ended strike involving 42,000 workers beginning May 14. If not resolved by July, it would carry into July as an ongoing stoppage. However, BLS counts workers “beginning in the period” - so this would count toward May, not July, unless it expands or a new stoppage phase begins in July.
The P(0) probability was too high: Given the hotel workers contract expires precisely on June 30, 2026, the probability of zero workers in July should be lower.
The BLS methodology counts stoppages “BEGINNING in the period”: The UC AFSCME strike beginning May 14 would count in May’s figures, not July. Only if it expands significantly or a new phase begins in July would it impact July numbers.
Hotel workers number clarification: The search results clarify it’s approximately 28,000 workers covered by the CBA (per Akerman LLP article), not 40,000. The union represents “nearly 40,000 non-managerial employees” overall across NYC metro, Capital Region, and NJ, but the IWA covers ~28,000.
Contract expiration ≠ strike: Many contracts are settled without strikes. The hotel industry faces “tremendous headwinds” which could push both sides toward resolution. However, the union’s aggressive posture (“We know what we need to win in 2026, and it’s a lot”) suggests elevated strike probability.
July-specific exposure is significant: The NYC hotel workers contract expires June 30/July 1 - this is a bullseye for July. If they strike, it likely begins July 1.
Writers Guild of America contract expires May 1, 2026: Could drag into July if negotiations extend like 2023 (148-day strike).
Multiple contracts with summer expirations: Rubber/tire (late July), IBEW Richmond (July), Kaiser NNU (August).
Given:
I should:
Updated prediction:
Percentile 0.1: 0 Percentile 1: 0.01 Percentile 5: 0.05 Percentile 10: 0.1 Percentile 20: 0.2 Percentile 30: 1000 Percentile 40: 3500 Percentile 50: 8000 Percentile 60: 16000 Percentile 70: 26000 Percentile 80: 38000 Percentile 90: 65000 Percentile 95: 110000 Percentile 99: 200000 Percentile 99.9: 400000
Looking at the critic’s feedback, I want to search for any updates on the specific labor disputes mentioned that could affect July 2026 forecasting.
This is very helpful new information! Let me also check on the LIRR situation:This is excellent updated information. Now I have a much clearer picture:
The critic raises several valid points that I should acknowledge:
Technical Issue with Low Percentiles: The critic is absolutely right that BLS only tracks stoppages with 1,000+ workers, so outcomes are either exactly 0 or ≥1,000. My original low percentile values (0.01, 0.15, etc.) were artifacts that don’t represent real possible outcomes.
P(0) Estimate Too Low: The research brief suggested 40-50% base rate for zero, but I originally implied only ~20%. However, I want to push back partially on this - current labor conditions are elevated, which justifies some reduction from the historical base rate.
New Evidence from Searches:
32BJ SEIU (NYC Building Workers): A tentative deal was reached on April 18, 2026 - avoiding what would have been a strike of 34,000 workers. This significantly reduces July 2026 risk from this source.
LIRR Workers: The strike deadline of May 16, 2026 looms for ~3,500 workers. Negotiations are stalled (no meeting since March 20), and both sides appear far apart. If this strikes in May, it may not affect July unless it begins in July specifically. However, LIRR workers are covered by the Railway Labor Act, making strikes difficult but possible.
TWU Local 100 (NYC Transit): Contract expires May 16, 2026 for 40,000+ members. Negotiations just began April 9, 2026. This is covered by the Taylor Law, which makes public employee strikes illegal in New York - historically, this significantly reduces strike probability (TWU Local 100 last struck in 2005).
Key Insights for July 2026:
Revised P(0) Estimate: Given:
I now estimate P(0) ≈ 33-38%, which is higher than my original ~20% but lower than the critic’s implied 40%+ given the still-elevated labor environment and pending LIRR situation.
Adjusting the Distribution:
Updated prediction: Percentile 0.1: 0 Percentile 1: 0.1 Percentile 5: 0.2 Percentile 10: 0.3 Percentile 20: 0.4 Percentile 30: 0.45 Percentile 40: 2000 Percentile 50: 4500 Percentile 60: 9000 Percentile 70: 17000 Percentile 80: 30000 Percentile 90: 52000 Percentile 95: 88000 Percentile 99: 165000 Percentile 99.9: 310000
Looking at the critic’s feedback, I need to address several key points:
Impossible values in lower percentiles: The critic is absolutely right. Values like 0.1, 0.15, 0.2 workers are impossible outcomes since BLS only tracks stoppages with ≥1,000 workers. The distribution must have a point mass at 0, then jump to ≥1,000.
Upper tail conservatism: Given SAG-AFTRA involved ~160,000 workers in July 2023, my 95th percentile at 100,000 was arguably too conservative. The elevated current labor tensions support a slightly fatter upper tail.
Bimodal nature: The critic correctly identifies that this distribution is essentially “nothing happens OR something significant happens.”
My fundamental assessment remains valid:
Proper handling of the discrete/continuous nature: I should use values that are essentially zero (0, 0.01, 0.05, etc.) for percentiles below the probability mass at 0, then let the interpolation cross 999.5 at approximately the 45th percentile.
Conditional median: If P(X=0) ≈ 45%, and conditional on at least one stoppage the median size is around 3,000-4,000 workers, the unconditional 50th percentile should be around 1,600-1,800 workers.
Upper tail calibration: Multiple major strikes coinciding (healthcare + transportation + entertainment) could push outcomes to 200,000-300,000 in extreme scenarios.
July-specific risk: The pending NYC transit contracts (May 16 expiry) could extend disputes into July, though resolution is also possible.
Updated prediction: Percentile 0.1: 0 Percentile 1: 0.01 Percentile 5: 0.05 Percentile 10: 0.1 Percentile 20: 0.2 Percentile 30: 0.3 Percentile 40: 0.45 Percentile 50: 1800 Percentile 60: 5000 Percentile 70: 12000 Percentile 80: 28000 Percentile 90: 58000 Percentile 95: 115000 Percentile 99: 230000 Percentile 99.9: 450000
The critic’s Round 5 analysis makes several compelling final points that warrant careful consideration.
“Main Character Syndrome” - This is an excellent meta-point. Both the critic and I have focused heavily on NYC hotel workers, but the actual July 2026 number could be determined by something entirely different—perhaps a hospital strike in the Midwest, or a delayed UC California action that we haven’t emphasized enough.
Timing Precision Risk - This is an important point I hadn’t fully considered. If the hotel contract expires June 30 (as the union website states), a strike beginning that day would count for June, not July. This timing risk meaningfully reduces the July-specific impact of the hotel situation.
Slightly Lower Median - With the timing risk and settlement dynamics, 10,000 is more appropriate than 10,500. The difference is small but defensible.
Upper Tail Adjustment for UC California - The AFSCME 3299 (40,000 workers) May strike could delay to July if negotiations extend, justifying slightly higher extreme percentiles.
P(0) at ~32% - This reflects the balance of education dormancy (pushing higher), hotel contract expiration (pushing lower), settlement dynamics (pushing higher), and timing risk (pushing higher). The current distribution captures this well.
Bimodal Structure - The jump from P50 (10,000) to P60 (23,000) appropriately captures the “either quiet month OR hotel strike scenario” dynamic.
Healthcare Sector Baseline - Ongoing healthcare militancy (Kaiser NorCal nurses approaching August expiration) provides baseline risk for moderate stoppages independent of the hotel situation.
Updated prediction:
Percentile 0.1: 0 Percentile 1: 0.1 Percentile 5: 0.2 Percentile 10: 0.4 Percentile 20: 0.47 Percentile 30: 0.49 Percentile 40: 4500 Percentile 50: 10000 Percentile 60: 23000 Percentile 70: 36000 Percentile 80: 55000 Percentile 90: 94000 Percentile 95: 140000 Percentile 99: 235000 Percentile 99.9: 430000
Looking at the Round 2 critique, I need to carefully weigh the points raised.
1. P30 → P40 Transition: The critic is correct that P30 = 0.49 (rather than 0.42) creates a cleaner representation of the zero-mass boundary. This ensures the value definitively rounds to 0 before jumping to realistic stoppage territory.
2. Conditional Distribution Height: Given elevated labor activity (2023-2025 all saw 30+ stoppages) and specific NYC disputes that could spill into July, a somewhat higher conditional distribution is warranted. The NYC transit contract (May 16) creates non-trivial July spillover risk.
P(0) Estimation: The critic suggests 25-28%, but I maintain ~30% is appropriate because:
I’ll compromise at ~30% (P30 = 0.49).
Conditional Median: The critic’s implied conditional median (~20,000) seems high. Historical data suggests average workers per stoppage is ~10,000, and median should be lower due to right-skew. I’ll target conditional median around 12,000-14,000 (unconditional P65).
The jump from P30 = 0.49 to P40 = 2,500 creates unavoidable interpolation artifacts in the impossible 1-999 range. This is a format limitation I acknowledge but cannot fully resolve.
Updated prediction: Percentile 0.1: 0.01 Percentile 1: 0.05 Percentile 5: 0.12 Percentile 10: 0.2 Percentile 20: 0.35 Percentile 30: 0.49 Percentile 40: 2500 Percentile 50: 6000 Percentile 60: 12000 Percentile 70: 20000 Percentile 80: 35000 Percentile 90: 62000 Percentile 95: 105000 Percentile 99: 210000 Percentile 99.9: 380000
Extracted Percentile Values: {0.1: 0, 1.0: 0, 5.0: 0, 10.0: 0, 15.0: 0, 20.0: 1000, 25.0: 1500, 30.0: 2500, 35.0: 3500, 40.0: 5000, 45.0: 6000, 46.0: 6300, 47.0: 6600, 48.0: 7000, 49.0: 7300, 50.0: 7700, 51.0: 8100, 52.0: 8500, 53.0: 9000, 54.0: 9500, 55.0: 10000, 60.0: 14000, 65.0: 19000, 70.0: 26000, 75.0: 35000, 80.0: 48000, 85.0: 65000, 90.0: 90000, 95.0: 155000, 99.0: 370000, 99.9: 650000}
Model’s Answer:
What I’m predicting: The BLS “Number of workers involved in work stoppages — Beginning in the period — Jul 2026” figure, measured in individual workers.
Recent monthly figures:
Annual context:
Upcoming labor landscape:
Why Low Outcome (0-5,000): Monthly data is extremely lumpy. With ~2.5 new major stoppages per month on average, some months see zero new stoppages beginning. No announced July actions as of today. Probability ~30-35%.
Why High Outcome (50,000+): A single large-scale stoppage (education, healthcare, public sector) can involve tens or hundreds of thousands of workers. Surprise events like failed last-minute negotiations are always possible. Elevated labor activity environment continues. Probability ~15-20%.
Assumption 1: No currently unannounced mega-stoppage (100k+ workers) begins in July. Probability of failure: ~8-10%. Impact: Would dramatically shift the distribution upward.
Assumption 2: The general labor environment in July 2026 remains similar to early 2026 (moderate stoppage activity). Probability of failure: ~15%. Could shift either direction if economic conditions deteriorate sharply or if political events trigger mass actions.
Assumption 3: BLS publishes data in time for resolution by August 10. Probability of failure: ~5%. BLS typically publishes with a lag; July data might not be available by August 10, creating resolution ambiguity.
The monthly distribution is heavily right-skewed. The median month likely sees 5,000-10,000 workers, while the mean is pulled higher (~20,000-25,000) by rare large events. With no specific July stoppages announced and 2+ months of lead time for surprises, I center my median around 7,000 with substantial uncertainty in both tails.
Percentile .1: 0 Percentile 1: 0 Percentile 5: 0 Percentile 10: 0 Percentile 15: 0 Percentile 20: 1000 Percentile 25: 1500 Percentile 30: 2500 Percentile 35: 3500 Percentile 40: 5000 Percentile 45: 6000 Percentile 46: 6300 Percentile 47: 6600 Percentile 48: 7000 Percentile 49: 7300 Percentile 50: 7700 Percentile 51: 8100 Percentile 52: 8500 Percentile 53: 9000 Percentile 54: 9500 Percentile 55: 10000 Percentile 60: 14000 Percentile 65: 19000 Percentile 70: 26000 Percentile 75: 35000 Percentile 80: 48000 Percentile 85: 65000 Percentile 90: 90000 Percentile 95: 155000 Percentile 99: 370000 Percentile 99.9: 650000
Forecast: median = 8886.7369
Forecast (percentiles, aggregated):
Per-model forecasts:
Rationale: Summary The baseline expectation for major work stoppages beginning in July sits around 6,800 workers, reflecting historical trends where a typical active month sees between 4,000 and 15,000 workers idled. However, the distribution of strike activity is heavily right-skewed. The data frequently features months with zero major stoppages, as seen in roughly 20-25% of recent July reports, which supports a lower-bound estimate of 0. Conversely, occasional massive strikes can inflate the figures dramatically. The labor environment heading into mid-2026 is highly active, continuing the momentum of 2025, which saw over 300,000 workers involved in major stoppages. Several exceptionally large union contracts are set to expire in late June and July 2026, most notably the SEIU Local 1000 contract covering 96,000 California state workers and the NYC Hotel Trades Council contract covering 30,000 workers. While the majority of contract disputes are settled without a strike, this clustering of large expirations creates significant upper-tail risk, driving the higher percentile estimates well above the historical median.
Strongest Arguments for Higher Values
Strongest Arguments for Lower Values
Key Uncertainties
Resolution will use the BLS monthly series for workers involved in major work stoppages beginning in July 2026. BLS defines major work stoppages as strikes or lockouts involving 1,000 or more workers and lasting at least one full shift; monthly data cover the full calendar month. BLS’s worker count includes workers directly involved and related idleness at directly involved establishments, but excludes secondary idleness. BLS also notes that monthly figures are preliminary until the annual release. (bls.gov)
The base rate argues for a usually modest July, but with a fat right tail. BLS reported 33 major stoppages beginning in 2023, 31 in 2024, and 30 in 2025; over 2016-2025 the average was 20.8 per year. Workers involved in annual starts were 271,500 in 2024 and 306,800 in 2025. For July specifically, BLS’s 2024 fact sheet shows 2 stoppages and 4.2 thousand workers in July 2024. From the BLS detailed monthly listing, I infer July 2025 summed to about 14.5 thousand workers: Philadelphia (9,000), Republic Services (1,300), Sacramento County (3,000), and Chukchansi Gold Resort and Casino (1,200). July 2023 was a major outlier at about 175 thousand workers, driven by Los Angeles/Orange County hotels (15,000) and SAG-AFTRA (160,000). July 2022 had no major starts in the detailed list, and July 2021 had one 1,400-worker start. So the recent July pattern is roughly 0, 1.4k, 0, 175k, 4.2k, 14.5k: usually low, occasionally enormous. (bls.gov)
Current 2026 data do not yet look like a 2023-style blowout year. The latest BLS page shows March 2026 at 4 major stoppages beginning and 10.0 thousand workers involved; an earlier cached BLS page showed February 2026 at 4 stoppages and 7.1 thousand workers. That is active, but not explosively so. (bls.gov)
What keeps the right tail alive is visible July-adjacent bargaining risk. SEIU Local 1000 says its current agreement preserves a path to bargain again in 2026 and keeps the return-to-office pause through July 1, 2026; Labor Notes reports the State of California contract covers 96,000 SEIU workers and expires June 30, 2026. Labor Notes also points to late-July master-contract expirations covering 4,000 Goodyear workers and 2,400 Bridgestone-Firestone workers, plus a July expiration for the New York City hotel master contract. On the offsetting side, UFCW says more than 12,000 Kaiser Permanente members ratified successor contracts through March 2030, which removes one otherwise plausible spring/summer 2026 healthcare flashpoint. (seiu1000.org)
My synthesis is therefore: a quiet July remains quite plausible; a low-to-mid five-figure month is the modal outcome; and one large public-sector, hotel, or industrial dispute could push the number toward roughly 96,000 or even above 100,000. I therefore center the forecast in the low tens of thousands, with the mode on the 4,000-8,000 portion of the client grid, an expected value around 16,000 workers, about 24% probability on zero, most probability mass below 20,000, and a meaningful but minority right tail above 50,000. Because the client requires 1,000-worker grid points while BLS often publishes 100-worker increments, the code maps each grid point to a 1,000-wide bucket centered on that value; 0 represents [0, 500), and the final bucket is the upper tail above 100,000.
As a resolution reminder, BLS counts major work stoppages that involve at least 1,000 workers and last at least one full shift; the monthly tables are preliminary until the annual release, and BLS notes that workers can be counted more than once if they participate in more than one stoppage during the reference period. (bls.gov)
The background labor environment is active, but not at an all-time extreme. BLS reported 30 major work stoppages beginning in 2025 involving 306,800 workers, versus an average of 17.8 stoppages beginning per year over 2006-2025. The latest BLS home-page monthly reading available as of April 27, 2026 showed 4 stoppages and 10.0 thousand workers beginning in March 2026. (bls.gov)
July specifically is usually a low-to-moderate month, but with a fat right tail. Recent July examples show that range clearly: the detailed listing implies only about 4,200 workers beginning in July 2024, while the BLS home page showed 15.5 thousand workers beginning in July 2025. But July 2023 was a true outlier: BLS’s detailed listing shows a 15,000-worker Los Angeles/Orange County hotel stoppage and a 160,000-worker SAG-AFTRA stoppage both beginning in July, implying about 175,000 workers that month. (bls.gov)
That 2023 example is why I reserve meaningful probability mass above 100,000. SAG-AFTRA says its current AMPTP contract expires on June 30, 2026, formal negotiations began on February 9, 2026, and the union represents about 160,000 members. The DGA is also in the same bargaining window: its current contracts expire on June 30, 2026, and its BA/FLTTA negotiations are scheduled to begin on May 11. July 2026 is therefore the natural month for a very large entertainment-industry stoppage if those talks fail. (sagaftra.org)
Against that upside tail, there are also meaningful settlement signals. The Writers Guild reached a tentative four-year agreement in early April 2026, and AP reported that members ratified it on April 24, 2026. I take that as evidence that studios and unions still have a strong incentive to avoid another 2023-style multi-union shutdown, even if it does not eliminate the risk. (apnews.com)
So my forecast is a mixture model. Most of the probability mass sits on ordinary July outcomes: no new major stoppage, one small stoppage, or a handful totaling roughly 1,000-20,000 workers. I then add a smaller sub-100,000 tail for a big municipal, healthcare, or public-sector dispute, and a distinct >100,000 tail for a SAG-AFTRA-style Hollywood strike. That produces a low median (around 4,000 workers) but a much higher mean because the upside tail is real. In the required grid, my single most likely exact bucket is 0 workers, but I assign substantially more combined probability to nonzero outcomes below 20,000 than to 0 alone, and I put 0.13 probability on the >100,000 bucket.
Net: I think “below 100,000” is clearly more likely than not, but not overwhelmingly so, because the June 30, 2026 Hollywood contract cliff creates a live path to a July result far above the client’s expected range. (sagaftra.org)
I forecast a distribution centered in the low five figures, with a mean around 15,000 workers. My modal region is the ordinary-July range of roughly 3,000 to 8,000 workers, but I include a meaningful bump around 28,000 to 32,000 workers because the New York City hotel contract cycle creates a concrete July-2026 strike risk. I also leave a small but real right tail for an unexpected national dispute. (hotelworkers.org)
First, I anchored on BLS methodology. BLS defines major work stoppages as strikes or lockouts involving 1,000 or more workers and lasting at least one full shift. The monthly ‘workers involved in all work stoppages beginning in the period’ measure sums workers in stoppages that begin in that month, excludes earlier stoppages that merely remain in effect, and is rounded to the nearest hundred. BLS also notes that monthly data are preliminary until the annual release. That matters because this question is specifically about new July 2026 stoppages, not all stoppages active in July. (bls.gov)
Second, I looked at the current labor backdrop. BLS reported 33 major work stoppages beginning in 2023 with 458,900 workers involved, 31 in 2024 with 271,500 workers, and 30 in 2025 with 306,800 workers. For 2026 so far, the BLS time series shows 3 stoppages and 47,700 workers in January, 4 stoppages and 7,100 workers in February, and 4 stoppages and 10,000 workers in March. So 2026 has started with an elevated count pace relative to the last two annual totals, which nudges me somewhat above a purely quiet-July baseline. (bls.gov)
Third, I used the recent July base rate. In the BLS July series for 2016-2025, the number of stoppages beginning in July was 1, 1, 2, 0, 0, 1, 0, 2, 2, and 4. The corresponding July worker totals were 2.6k, 1.2k, 4.2k, 0, 0, 1.4k, 0, 175.0k, 4.2k, and 14.5k. That means July is usually a small month or a zero month, but it occasionally produces a very large outlier. Using those BLS values, the recent-July average is 1.3 stoppages and 20.3k workers, but that worker average is heavily distorted by the July 2023 spike; excluding 2023, the average drops to about 3.1k workers. (data.bls.gov)
The detailed monthly listings reinforce that ordinary July months are usually not huge. BLS’s listing for July 2024 shows two stoppages totaling 4,200 workers: Rady Children’s Hospital at 1,600 workers and the video-game voice-actors strike at 2,600 workers. July 2025 totaled 14,500 workers across four stoppages: City of Philadelphia at 9,000, Republic Services at 1,300, Sacramento County at 3,000, and Chukchansi Gold Resort and Casino at 1,200. That is why my baseline component is zero-inflated and concentrated below about 15,000 workers rather than near the 2023 outlier. (bls.gov)
The biggest identifiable July-2026 upside risk is the Hotel and Gaming Trades Council’s Industry-Wide Agreement in New York City. HTC says that agreement expires in July 2026, and its history page says the IWA covers 28,000 members. HTC President Rich Maroko also wrote that the union is preparing members for the 2026 contract fight, explicitly discussing the need to be ready for a city-wide strike if necessary. At the same time, Maroko notes that similarly intense preparations ahead of the 2006 negotiations ended in a strong contract without an actual strike. So I treat this as a substantial July-specific risk, but still clearly below even odds. (hotelworkers.org)
Putting those pieces together, I used a four-part mixture model. (1) A 76% ‘ordinary July’ component captures the usual zero-to-low-five-figure outcome range suggested by most recent Julys. (2) A 17% ‘NYC hotel contract fight’ component centers around roughly 30,000 workers, reflecting the 28,000-member IWA plus some room for other same-month actions. (3) A 5% ‘other large surprise’ component centers around roughly 60,000 workers, representing the chance of some unrelated national or multi-site dispute. (4) A 2% extreme-tail component captures something more like a July 2023-style shock and generates most of the probability above 100,000. I then discretized to the client’s required 1,000-worker grid by assigning continuous probability to nearest-1,000 bins. (data.bls.gov)
Net: my median is in the single-digit thousands, my mean is around 15,000, I see the single most important non-baseline cluster around 28,000 to 32,000 workers, and I assign only a small tail to outcomes above 100,000.
I start from the BLS definition. For this question, the relevant series is the number of workers involved in major work stoppages “beginning in the period,” where major work stoppages are strikes or lockouts involving 1,000 or more workers and lasting at least one full shift. BLS counts workers directly involved or idled for related reasons at directly involved establishments, not broader secondary spillovers. The latest BLS pages show 4 stoppages and 7.1 thousand workers beginning in February 2026, and 4 stoppages and 10.0 thousand workers beginning in March 2026, which suggests 2026 has remained active but not yet at a 2023-style blowout pace. (bls.gov)
For base rates, BLS reported 30 major work stoppages beginning in 2025 and 306,800 workers involved. Over 2016-2025, BLS says the annual count averaged 20.8 stoppages. Recent annual worker totals were 458,900 in 2023, 271,500 in 2024, and 306,800 in 2025, so the post-2022 environment has clearly been more strike-prone than the pre-2022 lull, though still nowhere near the mid-20th-century historical extremes. In 2025, 17 of the 30 stoppages beginning in the year were in state and local government, versus 13 in private industry, which matters because public-sector disputes have recently been a larger share of the mix. (bls.gov)
July specifically is highly skewed. Official BLS sources imply July 2023 was about 175,000 workers beginning in the month (15,000 hotel workers plus 160,000 SAG-AFTRA), July 2024 was 4,200, and July 2025 was 14,500. That range tells me a “normal” July lately is often in the low tens of thousands, but a single national-scale dispute can dominate the month. BLS’s historical factsheet reinforces that tail risk: among the five largest stoppages since 1993 were actions involving 135,000 to 180,000 workers. (bls.gov)
The most important 2026 upside-tail risk I see is California state workers. SEIU Local 1000 says it represents about 96,000 workers, while CalHR’s bargaining summary for the covered units lists approximately 106,982 full-time equivalents. The current MOU runs through June 30, 2026, and the June 28, 2025 side letter explicitly preserves bargaining over the successor MOU; that same side letter also sets July 1, 2026 as the return-to-office reinstitution date unless changed. SEIU’s 2026 materials repeatedly frame 2026 as a major contract fight. That creates a real, though still minority, chance that July 2026 could include a very large public-sector stoppage near 100,000 workers. (seiu1000.org)
My central view is still below that tail. The broader BLS history since 1993 shows stoppages concentrated in manufacturing, health care and social assistance, educational services, transportation/warehousing, and public administration, and 2025’s public-sector tilt nudges me a bit upward for summer 2026 risk. But early-2026 BLS monthly numbers are not yet signaling another July 2023-type national labor shock as the base case. So I put most probability mass between about 8,000 and 30,000 workers, with a median around 15,000 and a mean around 22,600. I reserve a low-single-digit probability for an outcome above 100,000, mainly because one SEIU Local 1000-scale stoppage could do it, while also keeping substantial mass in the 30,000-70,000 range for a busy but non-mega month. Because your required output only allows exact 1,000-worker grid points while BLS can publish non-multiples such as 14,500, the code discretizes a continuous scenario model onto that grid using 500-worker half-width bins. (bls.gov)
BLS defines major work stoppages as strikes or lockouts involving 1,000 or more workers and lasting at least one shift, and its monthly workers-involved-beginning series is the target here. July history is lumpy: BLS’s July workers-beginning values for 2016-2025 were 2.6k, 1.2k, 4.2k, 0, 0, 1.4k, 0, 175k, 4.2k, and 14.5k; July stoppage counts were 1, 1, 2, 0, 0, 1, 0, 2, 2, and 4. That implies a roughly 20k ten-year July average, but only because of the 175k July 2023 outlier; the baseline July is much closer to low single-digit thousands. (bls.gov)
The 2026 backdrop is active but not yet dominated by giant start-month events. BLS shows 3, 4, and 4 stoppages beginning in January, February, and March 2026, involving 47.7k, 7.1k, and 10.0k workers respectively. BLS’s 2025 annual report also showed 30 stoppages and 306.8k workers, with state and local government accounting for more stoppages and far more idled workers than private industry, so public-sector and service-sector tail risk remains real. (data.bls.gov)
The main reason I am above the low historical July median is the calendar of June 30 and early-July contract exposure. The biggest identified July-tail risk is New York City hotels: HTC says its industry-wide agreement is set to expire in July 2026 and has been preparing members for a city-wide fight, while an employer-side January 2026 memo says the CBA covering about 28,000 employees expires at midnight on June 30, 2026 and that a work stoppage is possible. Another important tail is California SEIU Local 1000: bargaining began on April 22, 2026; the current agreement runs through June 30, 2026; Local 1000 says it represents about 95,000 employees; and the union explicitly says state workers have a legal right to strike and discusses strike leverage if negotiations are log-jammed by summer. UC is another summer risk cluster: UC says Teamsters Local 2010 covers 17,752 clerical workers with a March 31, 2026 expiry and UC-AFT covers about 5,265 lecturers with a June 30, 2026 expiry, although UC’s biggest academic-worker risk was reduced because UAW 4811 already ratified a new deal whose expiration was moved to December 31, 2029. (hotelworkers.org)
Putting that together, I expect most of the mass to stay in the 0-15k range, with a secondary hump around the high-20k range from a possible NYC hotel event and a long but much thinner right tail into 60k-100k from a California state-worker or overlapping public-sector scenario. My distribution therefore puts 24% on zero, centers the mean in the low-teens thousands, and leaves a small but non-trivial tail above 50k, including roughly a 1% chance of finishing above 100k. (data.bls.gov)