NHTSA is placing in the docket a Preliminary Economic Assessment (PEA) which analyzes the potential impact of the proposed new performance requirements and associated test procedures for advanced air bag systems. The Executive Summary of that document summarizes its conclusions as follows.
Compliance Scenarios
This analysis identified and analyzed three groups of possible compliance scenarios that combine the mandatory and optional test procedures for each risk group. Each scenario includes the three mandatory 5th percentile female dummy tests, as well as the existing 50th percentile male dummy frontal barrier tests with upgraded injury criteria. One scenario (Option #1) assumes that out-of-position children and driver requirements will be met with the out-of-position suppression test, while infant requirements will be met with the infant presence suppression test. A second scenario (Option #2) assumes that requirements for all three groups will be met with the low risk deployment test. A third scenario (Option #3) assumes that child and adult requirements are met with the dynamic out-of-position test, and the infant requirements are met with the infant presence suppression test.
Methodology
The analysis estimates the benefits and costs of incremental improvements in safety compared to two different baselines. The first is a baseline of pre-MY 1998 air bag vehicles. Tables E-1 and E-2 provide cost and benefits estimates assuming a pre-MY 1998 air bag vehicle baseline. The second baseline assumes that all vehicles are designed to the sled test and provide benefits in full frontal impacts (12 o'clock strikes), but no benefit in partial frontal impacts (10, 11, 1, and 2 o'clock strikes). Table E-3 provides costs and benefits assuming a baseline of vehicles designed to the sled test. Neither of these baselines reflect potential shifts in occupant demographics, driver/passenger behavior, belt use, child restraint use, or the percent of children sitting in the front right seat due to education efforts and labeling. The agency requests comments on alternative baselines, including ways to predict future changes in occupant behavior, and including the likely evolution of air bag designs in the absence of this rulemaking.
While primary and alternative injury criteria performance limits are proposed and analyzed in this assessment, only the primary proposal results are discussed in this executive summary.
Safety Impacts
Potential safety impacts of this proposal are dependent on the specific method chosen by manufacturers to meet the proposed test requirements. Some countermeasures reach a larger target population and potentially provide more benefits than others, although each might adequately meet test requirements. For example, a weight sensor could suppress the air bag up to its design limit for weight, but would not suppress the air bag for heavier occupants. Thus, in Table E-1, it is assumed that a 54 pound weight sensor would be utilized to meet the "Suppression When Presence" test with the 6 year-old dummy. While it could potentially save 102 children ages 1 to 12, it could not save all 129 children in that age category, because it is estimated that the remaining children will weigh more than 54 pounds. Multi-stage inflation systems are an example of a system that could potentially impact a wider range of injuries than do proximity sensors.
The ranges of potential safety impacts by test type are shown in Table E-1 and total fatality benefits for the three examined compliance options are shown in Table E-2. The estimated range of fatalities prevented from the three scenarios is 226-239 annually. Of these, 25 are in high speed tests and the remainder are in tests to minimize risks to out-of-position occupants. These estimated lives saved can also be broken into 167-175 passengers and 59-64 drivers. Injuries were not examined in this preliminary analysis because research to establish injury impacts has not been completed. However, the agency believes there will be significant injury reductions, particularly chest injuries.
| Table E-1 Estimated Target Population and Lives Saved Annually for the Primary Proposal Compared to Pre-MY 1998 Air Bags | |||||
|---|---|---|---|---|---|
| Tests | Drivers | RFCSS (passenger) | 1-12 Year Old Children (passenger) | Adult (passenger) | Total |
| Out-of-Position Target Population | 41 | 33 | 129 | 11 | 24 |
| Estimated Lives Saved by Different Tests (These are not additive) | |||||
| Suppression When Presence | NA | 33 | 102 | NA | 135 |
| Suppression When Out-of-Position | 41 | NP | 129 | 11 | 181 |
| Low Risk Deployment | 36-39 | 31-33 | 114-122 | 10 | 191-204 |
| Dynamic Out-Of-Position | 36-39 | NP | 114-122 | 10 | 160-171 |
| 25 mph Offset Barrier | 36-39 | 0 | 0 | 10 | 46-49 |
| In-Position Target Population | 6,778 | NP | NP | 1,501 | 8,279 |
| Estimated Lives Saved by Different Tests (These are additive) | |||||
| 30 MPH, Belted/Unbelted 50th Male | 11 | NP | NP | 0 | 11 |
| 30 MPH, Belted/Unbelted 5th Percentile Female | 5 | NP | NP | 1 | 6 |
| 25 MPH Offset Barrier | 7 | NP | NP | 1 | 8 |
NP: Not proposed test for this group.
Costs
Potential compliance costs for this proposal vary considerably and are dependent on the method chosen by manufacturers to comply. Methods such as modified fold patterns and inflator adjustments can be accomplished for little or no cost. More sophisticated solutions such as proximity sensors can increase costs significantly. Table E-2 lists the range of compliance costs for each compliance option. The range of potential costs for the compliance scenarios examined in this analysis is $22-$162. This amounts to a total potential annual cost of up to $2.5 billion, based on 15.5 million vehicle sales per year.
Property Damage Savings
Compliance methods that involve the use of suppression technology have the potential to produce significant property damage cost savings because they prevent air bags from deploying unnecessarily. This saves repair costs to replace the passenger side air bag, and frequently to replace windshields damaged by the air bag deployment. Property damage savings are shown in Table E-2. Property damage savings from these requirements could total up to $158 over the lifetime of an average vehicle. This amounts to a total potential cost savings of nearly $2.5 billion over the lifetime of a complete model year's fleet.
Net Cost Per Fatality Prevented
Table E-2 summarizes the cost per fatality prevented of each compliance option. Property damage savings have the potential to offset all, or nearly all of the cost of meeting this proposal. The maximum range of cost per fatality saved from the scenarios examined in this analysis is a savings of $9.4 million per fatality saved to a cost of $4.8 million per fatality saved. The range for passenger-side impacts is more favorable than for driver-side impacts. This is due to the potential property damage savings from suppressing air bags for children, and because there are far fewer out-of-position drivers at risk than there are passengers, particularly children. Passenger side costs vary from a savings of $14.7 million per fatality to a cost of $4.5 million per fatality. On the driver's side, costs range from zero to a cost of $21.2 million per fatality prevented.
| Table E-2 Summary of Costs and Benefits Compared to Pre-MY 1998 Air Bags | ||||||
|---|---|---|---|---|---|---|
| Cost Per Vehicle (1997 Dollars) |
Annual Total Costs (Billions) |
Annual Fatalities Prevented (after 7% discount) |
Lifetime Property Damage Savings Per-Vehicle | Net Cost (Net Savings) Per Vehicle |
Net Cost (Net Savings) Per Discounted Fatality Saved (Millions)** |
|
| Compliance Option #1 OOP Suppression*, Child Suppression |
$75 - $162 | $1.16-$2.51 | 239 (172) | $21 - $158 | $4 - $53 | $0.3 - $4.8M |
| Compliance Option #2 Low Risk Deployment |
$22 - $56 | $0.34-$0.86 | 226 - 233 (163 - 168) | $21 - $158 | $1 - $(102) | $(9.4) - $0.1 |
| Compliance Option #3 Dynamic OOP*, Child Suppression |
$24 - $162 | $0.37-$2.51 | 228 - 233 (165 - 168) | $21 - $158 | $2 - $4 | $0.2 - $0.4 |
*Note: OOP = out-of-position
All three options include offset barrier and frontal barrier tests.
** Net cost per discounted fatality saved is computed by taking the net cost per vehicle times 15.5 million vehicles divided by discounted fatalities prevented.
Sled Tests
Sled tests were temporarily allowed as an alternative method to certify compliance with FMVSS 208 in March 1997 in order to facilitate introduction of depowered air bags. A provision of the NHTSA Reauthorization Act (P.L. 105-178) provided that this method would remain in effect until changed by rule. This analysis thus addresses the relative merits of full frontal barrier tests and the sled test alternative. NHTSA is proposing to eliminate the sled test alternative because it is not representative of real world crashes that have the potential for serious injury or fatality, and it does not adequately test how well the vehicle and its restraint system protect outboard front seat occupants in those situations. Relatively modest changes have occurred thus far in air bag designs that use the sled test for compliance. However, NHTSA is concerned that potentially, air bag systems designed only to pass the sled test would expose occupants in higher speed crashes to significant increases in crash forces. For example, because the sled test is only a "12 o'clock" test, there is concern that it could lead to decreased air bag volume, which would provide less protection in frontal crashes at offset angles and to unbelted passengers in any frontal high speed crash. NHTSA examined air bag data supplied by nine auto manufacturers in response to an information request issued by the agency in December 1997. The agency found that of 42 passenger side model year 1998 systems examined, 10 had decreased air bag volume. Eight of these ten decreased the width of the air bag. This demonstrates that air bags designed to meet the sled test may provide protection to a smaller area of the occupant compartment, or in a narrower set of collision angles.
The effectiveness of air bags decreases as the crash moves further away from direct frontal impacts - 31 percent effective at 12 o'clock, 9 percent effective in 11 and 1 o'clock impacts and 5 percent effective in 10 and 2 o'clock impacts. If air bag designs provided no benefit in partial frontal impacts (10, 11, 1, and 2 o'clock), an estimated 319 lives would not be saved annually by air bags. In addition, the agency's analysis of limited test data of MY 1998 air bag vehicles versus pre-MY 1998 air bag vehicles estimated that 16 to 86 lives may not be saved in full frontal impacts by MY 1998 air bags that have been certified to the sled test. In total, 335 to 405 lives potentially would not be saved by vehicles designed to the sled test, rather than to the barrier test. Table E-3 shows that the net cost per fatality saved ranges from a savings of $3.4 million per fatality saved to a cost of $2.0 million per fatality saved.
In designing a low risk air bag, it will be more difficult for the manufacturers to meet all of the test conditions with an unbelted rigid barrier test than with a sled test. Many more sled tests than barrier tests can be run in a day and sled tests are less expensive to run than vehicle tests into a barrier. The development effort to design to the unbelted barrier test is more complex because many more factors have to be accounted for, including the angle test. The agency is not sure what would be the difference in vehicle costs between the two tests. If air bags are made smaller with the sled test, some minor savings in the air bag and sodium azide pellets would accrue. No additional cost has been added to Table E-3. However, since air-bag equipped vehicles have met the unbelted test in the past, there is little need to redesign air bags when suppression is the technology of choice.
