According to the National Motor Vehicle Crash Causation Survey (2005-07) between 28,500 and 59,100 crashes per year in the United States are due to a vehicle component failure or degradation.
Clearly, this is alarming, and I hate to be the bearer of bad news, but this estimation most likely falls short of the actual number. As with all surveys, there are official margins of error, but I suspect that this survey also contains errors of misclassification and/or assumption.
How Did the Government Gather Crash Data?
The National Highway Traffic Safety Administration (NHTSA) is an agency of the U.S. federal government, part of the Department of Transportation, and it focuses on "reducing deaths, injuries and economic losses from motor vehicle crashes."
Part of NHTSA's mission is to investigate potential improvements to the general public's vehicles. To do this, they collect research by conducting surveys and cross-referencing their databases, such as the National Automotive Sampling System's Crashworthiness Data System (NASS-CDS), which is also part of NHTSA.
Unfortunately, one of the drawbacks of these databases is that they do not examine pre-crash scenarios or the reasons behind the critical pre-crash events, which are "the immediate reason[s] for the critical event."
Essentially the NASS-CDS examines what happened in a crash, but not why it happened. As you can imagine, if you're the agency in charge of reducing highway fatalities, knowing what causes crashes is a big deal.
The National Motor Vehicle Crash Causation Survey: An Overview
That is where NHTSA's National Motor Vehicle Crash Causation Survey (NMVCCS) stepped in. Conducted from July 3, 2005, to December 31, 2007, the NMVCCS was "aimed at collecting on-scene information about the events and associated factors leading up to crashes" to better understand how we can prevent crashes and develop Crash Avoidance Technology (CAT).
NMVCCS did not focus on assigning fault. Rather, it examined the different facets and components affecting the drivers and vehicles involved in a crash. To do this, the researchers analyzed a crash's "pre-crash movement, critical pre-crash event, critical reason, and the associated factors."
The survey employed a weighted sample of 5,470 crashes (that resulted in a towed vehicle) to represent an estimated 2,189,000 crashes nationwide. Later, when synthesizing the data for publication, NHTSA focused its 2015 Crash Stats around the specific element of crash occurrence labeled the critical reason.
NHTSA reported that a crash's critical reason is "the immediate reason for the critical pre-crash event and is often the last failure in the causal chain of events leading up to the crash." The definition's key term is "causal." Basically, a critical reason is a carefully worded way to say what caused a crash without placing official fault.
What Causes Crashes (According to the NHTSA)
The NMVCCS researchers attributed critical reasons to 4 categories:
- Driver (94% of crashes),
- Environment (2% of crashes),
- Vehicle (2% of crashes),
- and Unknown Critical Errors (2% of crashes).
Anyone who has driven can generally understand why NMVCCS chose these categories. We know that drivers can make bad decisions and can cause collisions (driver), and we know that exterior factors can affect well-intentioned drivers (environment). Lastly, we know that a vehicle issue could be either improper maintenance or a defective product (vehicles).
In two percent of the examined crashes (or the weighted estimation of 44,000 crashes), "the critical reason was assigned to a vehicle component’s failure or degradation." This 2% has a plus or minus margin of error of 0.7%, giving us the range of 28,500 to 59,100 vehicle-defect crashes. Yet, this might not be the end of the story.
The unknown critical reasons category is intriguing. Let's take a look at this category to see if we can puzzle out something useful.
What About Those Pesky "Unknown" Critical Reasons?
The NMVCCS researchers had to add several caveats to their study methods and findings; the most problematic one being that the survey only used "on-scene information."
By only using data collected on-scene at a crash, the analyzed numbers were inherently limited and prevented researchers from considering "other internal vehicle-related problems." Thus, the NMVCCS could not synthesize in-depth details of any causes that were not visually apparent at the scene of the wreck.
Additionally, to get the visually apparent causes, the NMVCCS crash-specific variables required the researchers to rely on non-researcher reports, such as on-scene police officers. These non-researchers would not have had the tools at their disposal to properly assess any pre-crash causes. Nor would the non-researcher be able to properly assign any crash cause outside of driver error or an obvious vehicle malfunction.
With no forensic examination of these crashes, it’s a decent bet that internal vehicle-related problems caused many, if not most, of these "unknown critical reason" crashes.
For example, during NMVCCS, which was from July 3, 2005, to December 31, 2007, the infamous Chevy Cobalt ignition switch defect was in full swing, taking over a hundred lives and injuring hundreds more. The recall from General Motors didn't come until 2014, a decade after the defect was first discovered (and ignored) and seven years after this NMVCCS study. So it is likely that the researchers did not account for this defect, as they would have only found this causation through forensic examination. There's a decent chance, that researchers attributed these crashes to an unknown critical reason, instead of the vehicle.
Vehicle Defects Are Products Liability Cases
So while the survey reports the minimum number of vehicle-defect crashes as 28,500, the maximum number could encompass the vehicle-related crashes and the lion’s share of the unknown critical reason crashes. If we put those two categories together, the potential number of annual crashes caused by vehicle issues rises to between 41,634 and 133,526, or almost 1 in every 16 crashes.
I spoke with, personal injury and wrongful death attorney, Michael Grossman about the wide range of possible accidents caused by vehicle defects. He told me, "That doesn't come as a surprise. The government's data usually relies on police investigations. I've represented numerous clients in products liability cases over the years, and I can't recall a single time where the police blamed the crash on a defect."
Looking at the best crash data available shows how hard it is for researchers to pin down just how many vehicle defects result in wrecks. If trained researchers have this much trouble, just imagine how difficult it is for officers in the field to spot defects that cause crashes. If you don't think this matters, just look at how long it took investigators to identify the Chevy Cobalt defect and the more than 100 lives lost in the interim. Reliable vehicle defect data has real-world consequences.