Chrysler and the Jeep – Outlier?

So Chrysler has thrown down the gauntlet, and its claque has dutifully delivered its standing O. Atta boy, Chrysler, tell those regulators to stick it!

As usual, those opining about Chrysler’s public resistance to recalling the 1993 – 2004 Jeep Grand Cherokees and 2002 -2007 Libertys for defective fuel tanks haven’t a bloody clue. And so, as usual, The Safety Record Blog will put Chrysler’s shot across the bow into its proper context.

To recap: In November 2009, the Center for Auto Safety petitioned the National Highway Traffic Safety Administration to open an investigation into fuel-fed fires plaguing the early model Jeep Grand Cherokees, alleging that the plastic fuel tank’s placement behind the rear axle and below the rear bumper, and the lack of adequate shielding made it more vulnerable to rupture or leakage from rear-impacts and in rollovers. According to Fatality Analysis Reporting System (FARS) data, this design resulted in 172 fatal fire crashes with 254 fatalities, CAS said. The agency granted the CAS petition in August 2010, and opened a Preliminary Evaluation. In June 2012, ODI bumped up the investigation to an Engineering Analysis. Two weeks ago, NHTSA announced that it had requested that Chrysler recall the 1993-2004 Jeep Grand Cherokee and 2002 -2007 Liberty and was ready to go to an Initial Decision hearing if Chrysler refused.

So far, Chrysler has refused. Its preliminary defense was laid out in a “White Paper” (see NHTSA Drops Hammer on Chrysler Jeeps) Yes, it was written on virtual white paper. But we expected something a little more detailed than a three-page press release and a chart with writing so tiny, one needs to blow it up 500 percent to read it. Its basic argument is, and has always been, this: The Jeep Grand Cherokee and the Jeep Liberty met the Federal Motor Vehicle Safety Standard (FMVSS) 301 fuel tank integrity at the time, and the Jeep Grand Cherokee and Liberty, statistically are not outliers for rear-impact fuel-fed fires. (Chrysler has until next Tuesday to file its official response.) 

Sounds reasonable, no? Let’s unpack it.

Federal Motor Vehicle Safety Standards

The first thing a manufacturer says after announcing the launch of a design-related recall in the wake of a NHTSA investigation is: “Our product meets federal safety standards.” (Actually, the first thing they say is: “Oh *!@#.”) The public – and here we include all manner of journalistic types who put this in their stories, without understanding what an impoverished statement this is – actually believes that this is some sort of safety guarantee.

A Federal Motor Vehicle Safety Standard is a minimum standard – not state-of-the-art, not average – minimum, the lowest you can go and still sell the vehicle in the U.S. The most important part of the standard is the compliance test. During rulemaking, manufacturers argue long and hard about minute facets of compliance tests, because that’s where the standard counts. Most automakers will build in a compliance margin to ensure it makes the grade. So, for example a manufacturer may build a roof structure 20 percent stronger than required to make up for variations in the vehicle and the tests, to make sure it complies, and maybe even because they are concerned about occupants getting their heads crushed in a rollover. Manufacturers self-certify compliance – they do not file these tests with NHTSA – nor are they required to. NHTSA runs its own compliance tests on a select number of vehicles each year as a spot-check.  

Does the test measure what happens in real-world crashes or otherwise perform in the field? Maybe. Maybe not. Typically, there is a long lag time from when a standard is promulgated – decades or so – and when it is upgraded. So, many of these minimum standards were written decades ago without many significant amendments for a very long time. I think we can all agree that automotive research, testing and technology have not stood still in the same way. Still feeling confident about the protection those standards provide?

For those in the back of the room who were fooling around while the teacher was talking, consider this: even meeting minimum standards is not a defense against a defect. This principle is written in The Motor Vehicle Safety Act of 1966 – the godhead from which all automotive regulation emanates:

“‘Motor vehicle safety standard’ means a minimum standard for motor vehicle or motor vehicle equipment performance. ‘Motor vehicle safety’ means the performance of a motor vehicle or motor vehicle equipment in a way that protects the public against unreasonable risk of accidents occurring because of the design, construction, or performance of a motor vehicle, and against unreasonable risk of death or injury in an accident, and includes nonoperational safety of a motor vehicle.”

This concept is further enshrined the United States Code addressing strict liability as it relates to vehicle design:

“Because collisions between automobiles are so frequent and common that they must be considered an unavoidable incident of the normal and proper use of automobiles, an automobile manufacturer may be held strictly liable for a defective design which produce injuries.”

It has been NHTSA’s stance that meeting a minimum standard does not necessarily rule out a defect finding, because standards don’t measure everything, in all circumstances. In fact, the agency tends to measure a narrow band of performance as dictated by the compliance test.  And, meeting a minimum standard doesn’t necessarily mean that a product is safe – in the way the public tends to understand it. The agency has reiterated this point to manufacturers and other entities numerous times.

For example, in 1999, then-NHTSA Administrator Dr. Richard Martinez scolded child seat manufacturers for their cavalier attitude toward the safety of their products. In a letter, Martinez said “With the safety of our Nation’s children at issue, mere compliance with the minimum requirements of the standard is not enough; minimum standards should not be the most in safety design that manufacturers provide, When products are engineered with narrow compliance margins, the level of safety risk increases, even if the product is in technical compliance with the minimum standard. Only consumers don’t know that the phrase ‘meets or exceeds federal safety standards’ is only a faint endorsement of a component’s safety.”

In its request letter to Chrysler, NHTSA reminded the automaker that it could not hide its defect behind the meets-federal-safety-standards fig leaf:

“Chrysler maintains, and we do not dispute, that the Grand Cherokee and Liberty vehicles complied with the requirements of Federal motor vehicle safety standard (FMVSS) No. 301, Fuel System Integrity, that was applicable when the vehicles were manufactured. As NHTSA has noted in the past, a federal motor vehicle safety standard is a “minimum standard for a motor vehicle … performance.” 49 US c. 30102(a)(9). The existence of a minimum standard does not require NHTSA to ignore deadly problems. Viewed another way, a FMVSS does not preclude a finding of a safety related defect in a vehicle when supported by the evidence.”

That’s federal safety standards in general. What about FMVSS 301 in specific? One of the original 23 safety standards proposed in 1966, Fuel System Integrity was established in 1973. The compliance test consisted of 30 mph flat, rigid barrier crash test. The fuel spillage could not exceed one ounce from impact until the vehicle came to stop, or one ounce per minute in the 15-minute period after the crash. It was not significantly upgraded until 2003,when the compliance test was replaced with a high-speed, offset rear impact crash test at 50-mph, in which a 3,015 pound moving deformable barrier (MDB) which would contact 70 percent of the test vehicle’s rear-end.  (Automakers had to have 40 percent of their fleets comply by 2007, with full compliance with the new rule by 2009.)  And yet, this more stringent test doesn’t capture some aspects of the real-world picture – such as the underride that occurs when vehicles brake and nose downward beneath the bumper of the struck vehicle.

In the 30 years between, the agency evaluated the effectiveness of the standard, and found out that it was fairly useless. A 1991 report, “Motor Vehicle Fires in Traffic Crashes and the Effects of the Fuel System Integrity Standard,” concluded that while FMVSS 301 reduced post-crash fires by 14 percent for passenger cars; it did not produce a statistically significant reduction in fire-related injuries or deaths.  The study also found that fires in fatal car crashes had increased due to the aging fleet, and for light trucks, the evaluation found that implementation of the  standard had no effect at all.  The study also concluded that there were not enough data to conclude that lives had been saved and suggested the speed threshold in the standard may be too low. A 1998 NHTSA analysis of fire-related fatal crashes, A Case Study of 214 Fatal Crashes Involving Fire, showed that a burn-related fatality was more likely to occur in a rear impact – 46 percent of the fatalities occurred in rear impacts and although the majority of fires occurred in frontal crashes (58 percent).

Those studies show that rear-impacts are likely to produce fires with deadly consequences for occupants, and the old standard didn’t do much to mitigate it. Automakers, including Chrysler, knew all of this. As a matter of design, manufacturers have known since the 1960s that rear located fuel tanks were more vulnerable and over-the-axle tanks were recommended.  The exploding Pintos of the 1970s brought into public view the concept that fuel tanks behind the rear axle were a really bad idea. Again, as NHTSA pointed out in its request letter, after that debacle, automakers began locating gas tanks in less vulnerable locations. Then, it pulled up a bit of company history for Chrysler – a 1978 memo noting the location of the Horizon, Omni and K-Car fuel tanks, below the rear seat: “This location provides the protection of all the structure behind the rear wheels – as well as the rear wheels themselves – to protect the tank from being damaged in a collision.”  Some manufacturers have included fuel tank location and safety in their marketing like GM’s 1994 Cadillac ad that depicted the tank with this caption: “Fuel tank is located forward of the rear wheels and suspension for added security in a collision.”

An Outlier?

Now for Chrysler’s statistical wizardry. NHTSA claims that the 1993 – 2004 Jeep Grand Cherokee and 2002 – 2007 Liberty are outliers in rear-impact fire-related deaths. Chrysler claims they are not outliers. Who’s right? Everything depends on how the data is sliced.

First, take note that Chrysler’s “White Paper,” comparison chart differs from the statistical comparisons it made to NHTSA during the investigation. In the latter case, Chrysler compared the rear-impact fire-related fatality rates of the Jeep Grand Cherokee to peer vehicles such as the Toyota 4Runner, Ford Explorer, Nissan Pathfinder, Chevrolet Blazer, Mitsubishi Montero, Isuzu Rodeo, and Isuzu Trooper. Chrysler’s “White Paper” changes the comparable group to a different universe of vehicles in which rear-impact fire-related fatality rates includes “comparables” such as the Chevrolet Chevette, Corvette, Mazda RX7, and Toyota MR2, all sports cars, among others. 

The model with the highest rate of fires and fatalities in rear-impact crashes on Chrysler’s new chart of “comparable” vehicles is the 1995 – 1999 Chrysler AMC Talon/Mitsubishi Eclipse AWD.  Chrysler claims that neither this “comparable” vehicle nor any of the other models more likely than the Jeep Grand Cherokee to be involved in fire related rear impact fires were recalled.  Not so.  Chrysler and Mitsubishi recalled some 1995 – 1996 Talon / Eclipse AWD models.  For what?  A fuel leak from the fuel pump or gauge unit gasket. (Recall 96V158).       

NHTSA looked at a series of rear impact crashes, in terms of fatalities, fatalities with fire and fuel leaks and calculated comparative rates among other peer SUVs, using millions of registered vehicle years. Its analysis showed that the “MY 2002- 2007 Jeep Liberty and the MY 1993-2004 Grand Cherokee performed poorly when compared to all but one of the MY 1993-2007 peer vehicles, especially in terms of fatalities, fires without fatalities, and fuel leaks in rear end impacts and crashes.

The rates calculated by NHTSA and Chrysler basically represent a fraction, with the number of fire-related deaths as the numerator and some measure of exposure, like millions of registered vehicle years, as the denominator. Yet these rates sweep up the effects of a lot of variables besides the location of the gas tank. The characteristics of the driver, where it is driven, and how the vehicle is used could be relevant to how frequently the vehicle winds up in a fiery rear-impact crash to begin with.

Then there is the issue of population size. For vehicles representing small fleets, every incident raises up the rate proportionately more than in large fleets. For example, if we flip a coin a million times, the rate that heads will come up is likely to be 50 percent. If we only flip it once, and it lands on heads, the rate is 100 percent. Small populations don’t distort the rate, but they do distort the comparison.

If one scrunches up one’s eyes, and uses a magnifying glass to read the left side of Chrysler’s “White Paper” chart, you will find a lot of sports cars, and vehicles that are likely to have small fleets. Is a Jeep Grand Cherokee really comparable to a Corvette and a Talon? Are Corvettes likely to have high rates of fiery rear-impact crashes, because they are driven by risk-taking males at a high rate of speed, who tend to lose control, spin around and hit something at the rear? These are not trivial considerations when making comparisons.

NHTSA’s comparison group – popular SUVs with the same basic characteristics as the Jeep Grand Cherokee, minus the unfortunate gas tank location – appears to make the better comparison

Randy Whitfield of Quality Control Systems Corp. who regularly analyzes NHTSA data says that the best and cleanest comparison of rear impact fire-related fatalities as they relate to gas tank location would be between the rates for 1993-2004 Jeep Grand Cherokee and the Jeep Grand Cherokee after 2004, when Chrysler relocated the fuel tank. 

“[Chrysler’s] white paper is all about justifying the Jeep Grand Cherokee and their field experience through statistics,” he says. “The big question is whether fire deaths in the 1993 – 2004 Jeep Grand Cherokee and 2002 – 2007 Jeep Liberty could have been reduced by a more appropriate design. The statistical significance of differences in rates presented by Chrysler tells us nothing directly about whether the rates reflect reasonable choices in design.” 

Acceptable fire death rates can’t be set by the vehicles with the worst records, he adds. “If it were otherwise, then manufacturers would be free to use more and more risky designs as long as the resulting death rates were not the worst.  Of course, such a policy could only cause more and more fire-related vehicle deaths. This has been known for decades. ”