← The Crash Report
By The Numbers

741,195 Vehicles Have a Known Rollaway Defect. The Fix Arrives in 10 Months.

On June 30, Ford told NHTSA that 741,195 of its SUVs and trucks have a transmission defect that can damage the parking system and let them roll away in Park.[1] Ford also told NHTSA the fix won't be ready until April 2027. That leaves a ten-month window where three-quarters of a million vehicles are driving around with a known, manufacturer-admitted rollaway risk and zero available remedy.

~8.3 billion
Estimated miles these vehicles will accumulate during the 10-month remedy gap

The arithmetic is grim. The recalled vehicles are 2018–2021 Expeditions, Explorers, F-150s, Navigators, and Aviators. SUVs and trucks in this age bracket average roughly 13,500 miles per year according to NHTS survey data. Ten months of that is about 11,250 miles per vehicle. Multiply across the full recall population and you get approximately 8.34 billion vehicle-miles traveled by known-defective machines before the first wrench turns.[2]

But the miles-during-gap number is only the opening act. NHTSA's own recall completion rate data paints the longer picture, and it is considerably worse. Vehicles under five years old see an 83% completion rate. Drop into the 5–10 year bracket, which is exactly where every single one of these recalled Ford and Lincoln models now sits, and the rate collapses to roughly 60%.[3] Run the math: 741,195 vehicles at 60% completion means approximately 296,478 will statistically never see a repair bay. Not during the gap. Not after. Not ever.

~296,000
Recalled vehicles projected to never receive the remedy, based on NHTSA's age-adjusted completion rates

Those 296,000 permanently unfixed vehicles each accumulate roughly 13,500 miles per year. That is 4 billion miles annually of rollaway-risk driving extending indefinitely into the future, long after Ford closes out its paperwork and NHTSA files its completion report.

The defect itself is mechanical and somewhat counterintuitive. A valve body separator plate inside the 10-speed automatic restricts hydraulic fluid flow to the park valve. Under certain shift sequences, this causes the parking pawl to engage while the vehicle is still moving, producing an unintended braking event that damages the park system's internal components. Once damaged, the transmission can no longer reliably hold the vehicle stationary in Park.[4] Ford documented 285 reports of potential vehicle movement, 24 property damage allegations, and 9 injuries, two of which the company classified as emotional rather than physical.

Ford's interim notification letters, which started mailing in August, essentially tell owners: your vehicle has a confirmed defect that may let it roll away, and there is nothing we can offer you right now except the advice to engage your parking brake every time you stop.[5] Consumer Reports and The Drive both recommended the same workaround. What neither mentioned is that the parking brake only addresses the rollaway component. The other half of the defect, the random unintended braking while driving caused by the parking pawl engaging mid-shift, has no driver workaround whatsoever. You cannot prevent a hydraulic component from malfunctioning through careful parking habits.

The structural problem

Federal law requires manufacturers to notify NHTSA and vehicle owners as soon as a safety defect is confirmed, regardless of whether a remedy exists. The logic is sound: owners deserve to know about risks immediately. The unintended consequence is a system that generates warnings without solutions, and the data shows that the warning-without-solution interval has become structurally embedded in how recalls work. J.D. Power's analysis of 2013–2015 recall data found that 45 million recalled vehicles went unrepaired during that three-year period, with recalls exceeding one million vehicles achieving only a 49% completion rate compared to 67% for recalls under 10,000 units.[6]

In 2021 alone, NHTSA administered over 1,000 safety recalls affecting approximately 35 million vehicles, yet only 52% were remedied.[7] The average age of vehicles on U.S. roads has climbed to 12.6 years, pushing an ever-larger share of the fleet past the age threshold where completion rates functionally collapse.

The strongest counterargument

Ford's filing notes that the electronic parking brake automatically engages as a backup if the transmission range sensor fails to detect a Park shift. That is a meaningful safety layer. The 285 movement reports out of 741,195 vehicles represent a 0.038% incident rate, which by fleet-scale standards is low. And the ten-month timeline may simply reflect responsible engineering: developing a robust software calibration for a hydraulically complex 10-speed transmission across five vehicle platforms and four model years is not trivial work. A rushed fix that introduces new failure modes would be worse than a delayed one that works.

What the numbers actually say

The counterargument is reasonable. It is also irrelevant to the structural point. Whether this particular recall takes ten months or three, the system permits and routinely produces gaps between confirmed defect and available remedy, and those gaps compound with age-adjusted non-completion into a permanent residual risk pool. The 296,000 number is not a worst case. It is the statistical baseline, derived from NHTSA's own reporting to Congress. Ford's actual completion rate for this specific recall may be higher or lower, depending on the severity perception among owners and the effectiveness of dealer outreach. But the system-level pattern, where vehicle age drives non-completion more reliably than defect severity, is well established across decades of data.

Limitations

The 60% completion rate is an aggregate figure across all manufacturers, all defect types, and all recall sizes in the 5–10-year vehicle age bracket. Ford's brand-specific rate may differ. Rollaway defects may motivate more owners to seek repair than, say, a seat belt retractor issue, pushing actual completion above the baseline. The VMT estimates use NHTS national averages rather than model-specific odometer data, introducing uncertainty of roughly ±15% for individual models. And critically, we are projecting forward from historical rates. The actual completion picture for recall 26S48 will not be knowable for at least two years after the remedy becomes available.

What to do

If you own a 2018–2021 Expedition or Navigator, a 2020–2021 Explorer or Aviator, or a 2021 F-150: check your VIN at nhtsa.gov/recalls. Use your parking brake every single time you park, regardless of grade. Watch for the wrench warning light on your instrument cluster. And when the remedy becomes available in spring 2027, get it done. You have about a 60% chance of being the kind of owner who does. The data says the other 40% will read this sentence, nod, and never make the appointment.

Sources & References

  1. NHTSA, Recall 26S48, Ford Motor Company, June 30, 2026. 741,195 vehicles. nhtsa.gov/recalls
  2. National Household Travel Survey (NHTS), average annual VMT by vehicle type. nhts.ornl.gov
  3. NHTSA, Report to Congress: Vehicle Safety Recall Completion Rates. Age-adjusted rates: <5 yrs 83%, 5–10 yrs 60%, 11–15 yrs 52%, 16–20 yrs 38%. nhtsa.gov
  4. Consumer Reports, “Ford and Lincoln Recall Trucks and SUVs to Fix Rollaway Risk,” July 1, 2026. consumerreports.org
  5. Ford Motor Company recall acknowledgment: interim owner notification letters mailed August 2026; remedy owner notification expected April 2027. Per NHTSA filing and Fox Business reporting, July 1, 2026.
  6. J.D. Power SafetyIQ Analysis, 2016. 45 million vehicles recalled 2013–2015 went unrepaired; recalls >1M vehicles: 49% completion vs. 67% for <10K. thetruthaboutcars.com
  7. National Safety Council, Vehicle Safety Recalls brief. 2021: >1,000 recalls, ~35M vehicles, 52% remedied. nsc.org

Source: NHTSA recall database, NHTSA Report to Congress on Recall Completion Rates, NHTS VMT data, J.D. Power SafetyIQ, National Safety Council. Completion rate projections use aggregate NHTSA data, not Ford-specific rates. VMT estimates use national averages with ±15% model-level uncertainty. See methodology for caveats.