Longer Weeks in US Auto Plants for a Slightly Smaller Market

Posted by Saif Khan

The US new-vehicle market is down about 1.5% so far in 2026. Over the same stretch, production workers in motor vehicles and parts averaged 46.0 hours a week in October, up from 42.8 a year earlier. Sales are flat to slightly lower, and the work week is not.

Retrocausal’s November 2026 research, Longer weeks in US auto plants for a slightly smaller market, reads this gap as a sign that existing lines are absorbing more hours and more variety than their standards were written for. That is an inference, not a measured fact. The supporting data is still worth walking through, because it touches nearly every decision an industrial engineering or continuous improvement team makes in automotive manufacturing.

Infographic comparing US new-vehicle sales, down about 1.5% year to date, with auto production workweek rising from 42.8 hours in October 2025 to 46.0 hours in October 2026.
Sources: Cox Automotive; US Bureau of Labor Statistics. Compiled by Retrocausal, November 2026.

A Market That Is Not Growing

Cox Automotive’s initial estimate put the October 2026 US new-vehicle sales pace at 15.5 million units annualized. That is down from 16.0 million in September and 15.7 million in October 2025, and it is the slowest pace since January. Year-to-date sales sit about 1.5% below 2025. Actual October volume was 1.324 million units, 2.9% higher than a year earlier.

Inventory tells a similar story. Dealers ended October with 2.92 million new vehicles, 2.9% fewer than a year earlier, and 87 days’ supply against roughly 90. The more interesting number is the mix: model-year 2027 vehicles made up 32% of stock, where the next model year had reached 49% at the same point in 2025.

Retrocausal’s read is that the model-year changeover is running later and slower than last year, which means plants are building the outgoing model year for longer. Cox Automotive said in September that manufacturers appear to be limiting new-model production and putting profitability ahead of volume, which fits that picture.

Two other details from the sales data matter for planning. Through September, fleet sales were up 7% while retail sales to consumers were down 5%. And although Cox raised its full-year forecast to 16.1 million from 15.8 million, it also said downside risk is now higher.

More Hours and More Assembly Jobs

The BLS preliminary figures for October 2026 show production workers in US motor vehicles and parts averaging 46.0 hours a week, seasonally adjusted, against 42.8 in October 2025. That is 3.2 more hours, or about 7.5% more per worker, by Retrocausal’s arithmetic. Average hourly earnings were $34.76, up $2.21 over twelve months.

The assembly side shows a larger move. In September 2026, production workers in motor vehicle manufacturing averaged 48.3 hours a week, 3.8 more than a year earlier, which implies about 44.5 hours in September 2025. Assembly employment reached 298,800, up 10,100 over twelve months. Motor vehicle parts employment went the other way, falling 7,800 to 514,000.

So assembly plants added both people and hours in a year when the market did not grow, while parts plants shed jobs. Retrocausal’s conclusion is that more people on longer weeks means more shifts, more overtime and more stations running outside the week their standards assumed.

Why that matters on the floor

Time studies, standard work, line balancing and ergonomic assessments are all written against an assumed operating pattern. When the pattern shifts to longer weeks, added shifts and more overtime, the paperwork can stay unchanged while the reality it describes does not. The gap tends to show up late, as quality escapes, fatigue-related issues or a line that never quite hits its rate.

Closing that gap starts with seeing what stations actually do. Video-based floor analysis and video traceability are two ways teams compare the standard against current practice rather than against what the binder says.

Tariffs and Relocated Product

Tariffs are a direct cost for automakers and an indirect driver of plant workload. NPR reported on October 29, 2026 that one Detroit automaker put its 2026 tariff cost at about $2.5 billion, lower than it had forecast earlier and partly offset by refunds. The same report said materials, including computer chips, and diesel could cost that company up to $2 billion this year. Another Detroit automaker expects commodities, memory chips and logistics to be a headwind of about $2 billion year over year.

Steel, aluminum and copper imports have carried added tariffs since 2025. Kelley Blue Book put the average new-vehicle transaction price at just over $50,000 in September 2026, close to a record.

Capital is following the cost pressure into US plants:

  • Toyota said in July 2026 that it will spend $3.6 billion over four years to move Tacoma production from Tijuana, Mexico to San Antonio, Texas, with a new assembly line in a standalone building.
  • Honda announced $3 billion on November 9, 2026 for a central Ohio plant with capacity for 240,000 vehicles a year, due in 2029.

Retrocausal’s take is that tariffs pull volume into US plants over several years, and that until new capacity opens, existing lines absorb the extra hours, variants and relocated product. The new plants are years out, and the work is arriving now.

Relocated product arrives without validated method and time data unless someone builds it first. Teams that rely on predetermined time systems (PMTS) or digital work instructions can set up standards for a variant before it reaches the line, rather than after the first launch problems appear.

Hybrids on Combustion Lines

Hybrid share of US new-vehicle sales reached a record 16.3% in the second quarter of 2026, up from 13.0% a year earlier, with hybrid volume up 23%. Pure electric vehicles moved the other way: Cox Automotive counted 262,649 EV sales in the third quarter, down 39.5% from the record third quarter of 2025 and about 6.4% of new-vehicle sales. Electrified vehicles overall, including hybrids and plug-in hybrids, were nearly 25%.

Forecasts point the same direction. Mobility Global expects North American assembly of full hybrids to pass 3 million units in 2027, growing 25% a year, and to reach nearly one-third of light-vehicle assembly by 2030. Investment is already committed. Honda says its Ohio investments raise hybrid capacity there by 250% and that it is retooling an Ohio transmission plant for two hybrid systems. GM plans $1.25 billion to ready its Lansing Delta Township assembly plant for hybrid production.

The product gap behind the shift

The Murphy Automotive Product Pipeline (September 2026) counts more than $70 billion in EV write-downs and program cancellations. It describes model years 2026 to 2028 as a three-year product gap, with new product returning late in calendar 2028.

Retrocausal’s guess is that hybrids are being added to lines that already run combustion models. If so, station content and line balance change with no new platform launch to reset them. A dedicated launch normally brings fresh method development, time studies and balancing. Adding a hybrid variant to an existing mixed line can skip that, and the standards slowly drift from what the line actually does. Rebalancing a mixed-model line is a problem automatic line balancing is built for.

Fewer Recalls, Far More Vehicles

Sedgwick, using NHTSA data, counted 431 US recalls in the first half of 2026. That is down 6.5% from 461 a year earlier and the lowest six-month total in six years. The number of vehicles covered rose 111.6% to 23.48 million from 11.09 million, the highest half-year total in five years.

Electrical systems led all categories with 74 recalls and 6.23 million vehicles, the most in a decade. Back-over prevention systems were second by volume at 3.01 million vehicles, and powertrain was third with 23 recalls covering 2.44 million.

By Retrocausal’s arithmetic, that works out to about 54,500 vehicles per recall in the first half of 2026, against about 24,100 a year earlier. The authors infer that fewer recalls reaching far more vehicles is what high-volume shared components and long model runs produce when a defect escapes. The recall data does not prove that assembly standards are the cause, and the paper does not claim it does. It does show that the cost of any single escape is growing.

That raises the value of catching problems at the station. Step verification and AI-tracked work instructions with poka-yoke are aimed at mistakes made during assembly, and quality planning with P-FMEA helps teams decide where those checks belong before a variant launches.

What One Second Costs

Retrocausal includes a small illustrative calculation, which it labels as its own and not from any source. A final assembly line with a 60-second takt builds 60 vehicles an hour. If the average cycle drifts by one second, output falls to 59.02 an hour. Over two 8-hour shifts and 235 production days, that is about 3,698 vehicles a year.

Diagram of a 60-second takt line building 60 vehicles an hour, dropping to 59.02 with a one-second cycle drift, which equals about 3,698 fewer vehicles a year.
Illustrative calculation by Retrocausal, not from any source.

Few lines drift in one clean second, but the arithmetic shows why a takt or time standard that has not been confirmed against current operating conditions is a financial question as well as an engineering one. Teams that want to put numbers on this for their own lines can start with a time study and test the result against their takt, or estimate the return with the Kaizen Copilot ROI calculator.

Five Questions for Plant and Engineering Leaders

The paper closes with five questions. They are worth taking to a plant review as written:

  1. Takt and time standards. Sales are down about 1.5% year to date while production workers average 46.0 hours a week. When was each line’s takt and time standard last confirmed against the hours it actually runs?
  2. Extra hours. Assembly production workers averaged 48.3 hours a week in September 2026. Which stations absorb the extra hours, and do their standard work and ergonomic assessments assume that week?
  3. Model-year changeover. Model-year 2027 vehicles were 32% of inventory at the end of October, against 49% a year earlier. How long will the outgoing model year run on your lines, and how many engineering changes has it absorbed?
  4. Hybrid mix. Hybrid share reached 16.3% in the second quarter. What happens to station content and line balance when hybrid mix moves ten points on a line that also builds combustion models?
  5. Relocated and new product. Tariff-driven moves such as Tacoma, and new hybrid capacity, put more product into US plants. Is the method and time data for those variants ready before launch, and who keeps it current?

The full paper, including its sources and the one-second calculation, is available as a free PDF download.

Conclusion

The October 2026 data describes a US auto industry with a flat market and a busier shop floor. Assembly workers are putting in longer weeks, assembly headcount is rising while parts employment falls, the model-year changeover is lagging last year’s pace, hybrids are growing as a share of sales, and tariff-driven investment is moving product toward US plants years before the new capacity opens.

None of these trends is dramatic alone. Together, they suggest that line standards written for an earlier mix, schedule and work week are being asked to cover more than they were built for. The recall data adds weight to the concern, since defects that escape now reach far more vehicles per recall than they did a year ago.

For industrial engineering and continuous improvement teams, the practical task is to check standards against how lines actually run today, and to have method and time data ready for variants before they launch. If you want to see how that looks on a real line, you can schedule a demo with Retrocausal.

Source note: All figures come from Retrocausal’s November 2026 Automotive OEM Edition research paper, which cites Cox Automotive, the US Bureau of Labor Statistics, NPR, Manufacturing Dive, TechCrunch, Industrial Info Resources, Autobody News (on Sedgwick’s H1 2026 index), DBusiness (on Mobility Global’s forecast) and CDG News (on the Murphy Automotive Product Pipeline). Statements described as the authors’ reading are their inferences.

Related Blogs