X. Net Safety Effects and Costs of On-Off Switches

A. Effect of Turning Off Air Bags on the Performance of Some Seat Belts.

A number of industry commenters stated that deactivating air bags could result in substandard performance of the seat belts. Senator John McCain also sent NHTSA a letter requesting that the agency investigate this possibility.

A good general introduction to this issue appeared in an article on March 31 in the Kansas City Star:


The seat belts on some newer cars were designed to work with their air bags, automakers say. Alone, they will not protect a person in a serious crash as well as an older-style belt.


The newer belts allow a person to travel forward a few more inches than older belts, and when used in conjunction with air bags have some advantages, experts say. If the air bag is removed, however, the person faces a greater risk of head or chest injuries from hitting the steering wheel or dashboard.
In minor or moderately severe crashes, the redesign of the belt won't make a difference, auto and safety officials say. But in severe crashes, a person is more likely to travel forward far enough to hit the dashboard or steering wheel, sustaining head and chest injuries, they say.
When used with an air bag as designed, the newer belt has some definite advantages over the traditional one....


Because it is looser, it is less likely to break a rib or collarbone in a severe crash....That is particularly of concern for elderly people.


In older cars without air bags, the work of restraining an occupant falls solely on the belt...


The newer belt can...give way a little bit so that the air bag takes up some of the force of the crash and spreads it out over a broader section of your body...The result: fewer belt injuries.


Seat belts are required to meet minimum performance requirements in Standard No. 209, "Seat belt assemblies," and seat belt anchorages in vehicles are required to meet minimum performance requirements in Standard No. 210, "Seat belt anchorages." However, dynamically tested belts (automatic belts or manual belts with air bags) do not have to meet the requirement of Standard No. 209 that places a maximum of 30 percent on the amount of permitted webbing elongation. In addition, the anchorages for dynamically tested belts do not have to meet the anchorage location requirements of Standard No. 210. These requirements are not necessary for belts which are dynamically-tested, because the dynamic test ensures that the system works to protect the occupant from the type of injuries these requirements are designed to prevent. The elongation requirements also do not apply to belts that are equipped with "load limiters" and that are installed at a seating position with an air bag. A load limiter is a component of a seat belt system used to limit the levels of forces transferred to an occupant restrained by the belt during a crash. In very severe crashes, the forces in the seat belt system may rise above levels considered safe. If a belt system has a load limiter, parts in the system deform so that the belt forces transferred to the occupant do not rise above a predetermined maximum level. There are different designs of load limiters, ranging from simple folds stitched into the seat belt webbing that are designed to tear under a certain load, to more complex mechanical systems, some of which play out a small amount of additional webbing at incremental increases in load levels. The exclusion from the elongation requirements does not unnecessarily prevent manufacturers from using a design for these devices that operates by affecting the length of the webbing.

The exclusion from the elongation requirement is not likely to significantly affect the safety of the belt system. Although manufacturers may have designed belt systems in some air bag equipped vehicles with more "give" than those in non-air bag equipped vehicles, a 1991 NHTSA study showed that webbing in vehicles with air bags far exceeded Standard No. 209's requirements despite the exclusion from the elongation requirement. The study showed that maximum elongation, when tested according to the requirements of Standard No. 209, was 15 percent or less, or about half the permitted amount of elongation. NHTSA updated this study and again found that the maximum elongation was 15 percent or less.

Some manufacturers have, appropriately, been using the flexibility in Standard No. 209 to optimize their belt systems to work with air bags. Additional webbing elongation and load limiters would not normally be a problem in an air bag equipped vehicle, because the air bag would limit occupant excursion. This additional "give" in the seat belts is normally beneficial because it prevents the belt from causing injuries. However, some load limiters, those releasing a relatively large amount of additional webbing, could result in additional deaths and injuries if the air bags are turned off. Unfortunately, if the air bag can not function because it has been turned off, the "give" in these seat belts would increase the chance that occupants would hit their heads and upper bodies more easily on the steering wheel, the A-pillar, the windshield, or other hard parts of the vehicle interior, and suffer serious injury. In some cases, the only way to solve this problem might be by replacing the entire belt assembly.

Another type of safety device that could be affected by turning off the air bags is a seat belt pretensioner. These devices retract the seat belt webbing to remove slack almost instantly in a crash, thus enhancing the effectiveness of the seat belts by reducing the distance that the occupant might otherwise travel forward. Pretensioners are not powerful enough to pull the occupant back into the vehicle seat; they merely remove slack. Some seat belt pretensioners are triggered by the same sensor that actuates the air bag, and may be wired into the same circuit as the air bag. Therefore, unless on-off switches are designed correctly, turning off the air bag may also disable the seat belt pretensioners. Pretensioners are not required by NHTSA standards, but are an improvement added at the manufacturer's option. NHTSA is not aware of any belt systems with pretensioners that allow more slack to be introduced than is allowed by systems without pretensioners. However, the system is likely to be more effective if the pretensioner is not disconnected as a result of the installation and use of an on-off switch. To NHTSA's knowledge, all air bags in vehicles with pretensioners can be turned off without disabling the pretensioners.

The exclusion of air bag equipped vehicles from the requirements in Standard No. 210 may have also been used by manufacturers to optimize their seat belt anchorage locations for seat belts used in conjunction with air bags. The agency cannot quantify or even estimate the extent to which vehicle manufacturers have availed themselves of this opportunity. NHTSA's anchorage location requirements are intended to reduce the likelihood that occupants would "submarine" i.e., slide forward under the lap belt. Submarining would cause the seat belt loads to be transferred to an occupant up on the soft tissue of the abdomen instead of down on the pelvic bones, thereby increasing the likelihood of abdominal injury. The static test in Standard No. 210 is intended as a substitute for a dynamic test where the interaction between the occupant and the lap belt can be observed. Since manual belts used with air bags do not have to meet Standard No. 210's anchorage location requirements, manufacturers may have located the anchorage locations to optimize the interaction between the belt and the air bag in controlling the forward motion of the occupant. With the air bag turned off, the system as a whole will not operate as designed, and the chance of abdominal injuries could be increased.

A minority of vehicles have load limiters or seat belt pretensioners. Using information provided by manufacturers on the design of 1997 model year vehicles and sales numbers of 1996 vehicles, NHTSA estimates that vehicles with pretensioners will comprise only 5 percent of 1997 vehicle sales. Using the same information, NHTSA estimates that vehicles with load limiters comprise about 22 percent of 1997 model year sales. Very few models have both load limiters and pretensioners. Since the number of vehicles with these features has been increasing in recent years, the actual percentage of models with these features in the entire on-road vehicle fleet is lower than the percentage in 1997 model vehicles. Nonetheless, NHTSA expects vehicle manufacturers, dealers and repair businesses will take appropriate steps to inform consumers whether their vehicle is equipped with one of these devices and to advise them whether any modifications to the vehicle belt system should be made. The agency's information brochure advises vehicle owners to ask the manufacturer of their vehicle about this issue.

NHTSA agrees with the industry commenters that turning off the air bag could result in a seat belt system with less than optimal performance. Modern vehicle restraint systems are highly complex and integrated, with the seat belt and air bag components often designed to work together. The seat belt systems may not be designed to work alone. Taking out one component of the integrated system could result in reductions in performance. Because many of the features identified by NHTSA are designed to operate only when high loads are placed on the belt system, the presence of these features will be of no consequence in low severity crashes in which the air bag has been turned off, especially when a small/light weight person is using the belt. However, those features will be consequential in a more severe crash. In such a crash, the belts will not provide their full benefits for a vehicle occupant if that person's air bag is turned off.

B. Net Safety Effects and Costs.

People not in any of the four risk groups specified in this final rule will be worse off if they turn off their air bag. These people include the vast majority of teenagers and adults, including older drivers. By turning off their air bags, they will increase their chance of death or serious injury in moderate to serious crashes. Even belted occupants and the vast majority of short occupants will increase their risk of serious or fatal head, neck or chest injury if they turn off their air bags.

The net safety effects of retrofit on-off switch use will depend in part upon what proportion of the switch users are people at risk. Among persons in risk groups, the net safety effect of use of the on-off switch will depend on the whether that group is, on balance, benefited or harmed by air bags. For a group, like infants, which has had members fatally injured, but not saved, by air bags, use of the on-off switch to turn off passenger air bags will produce a net positive safety effect for the group. However, for other groups, use of the on-off switch to turn off driver air bags could have a net negative safety effect for the group.

Survey data provided by commenters suggest that many more people want on-off switches than could possibly benefit from them. As suggested above, the agency believes that this is because people tend to hear more about, and be more reactive to, the small number of fatalities from air bags than the large number of lives saved by air bags. The January 1997 survey provided by IIHS suggested that 30 percent of respondents were generally interested in on-off switches for the driver air bag, and 67 percent in on-off switches for the passenger air bag. Several commenters suggested that widespread availability of on-off switches would raise the possibility of what they termed "misuse," i.e., use of on-off switches by persons who are not at risk and who are clearly better off with their air bag left on. If this were to occur, it could result in a negative effect on safety. However, to the extent that the reported interest in on-off switches simply reflected a desire to make it possible to turn off an air bag should a person at risk ever be carried, then the likelihood of use by persons not at risk would be smaller.

As previously noted, the more recent IIHS survey, conducted in August, indicates that the general interest in on-off switches for passenger air bags has declined considerably since January. According to the new survey, 26 percent of respondents expressed a general interest in passenger air bag switches. General interest in driver air bag on-off switches was essentially unchanged, with 27 percent of respondents expressing an interest in those switches. The new survey also showed that interest in on-off switches declined after the respondents were informed about matters such as air bag benefits, steps for reducing risk and the cost of switches. The figure for passenger air bags dropped from 26 percent to 16 percent and the figure for driver air bags dropped from 27 percent to 12 percent.

To minimize the possibility of adverse safety consequences, persons who wish to apply for retrofit on-off switches must certify that they have read a NHTSA information brochure that explains the benefits and risks related to air bags to ensure that they make informed decisions both with respect to obtaining, and then using, an on-off switch. The brochure identifies which groups may be at risk, and which are not. More important, persons interested in on-off switches must certify that they or a user of the seating position in question meets the criteria for one of the relevant risk groups. Limiting eligibility for on-off switches to vehicle owners who are able to certify risk group membership should minimize the possibility that persons not in a risk group will have an opportunity to use a on-off switch to turn off their air bag and reduce the possibility that the switch will be used improperly. Finally, owners must submit their request to the agency for approval.

Given the large numbers of lives currently being saved by air bags and the very small chance of a fatality due to an air bag, and notwithstanding the limitation on eligibility for a on-off switch, NHTSA recognizes the possibility that authorizing the installation of retrofit on-off switches could result in a net loss of life. The agency has analyzed these adverse effects in its Final Regulatory Evaluation (see summary below). NHTSA notes that to the extent such a loss occurs, it would be the unfortunate result of several readily avoidable events: the incorrect certification of risk group membership, the use of on-off switches by persons who are not members of risk groups, and the failure to use seat belts and/or child restraints properly and to take other readily available precautionary measures.

NHTSA is issuing this final rule, notwithstanding its potential to reduce the number of lives saved by air bags, because the agency believes that it must consider both the short-run and long-run implications of this rulemaking on safety. Ultimately, the continued availability and use of any safety device, whether provided voluntarily by manufacturers or pursuant to a regulation, is dependent on public acceptability. The agency believes that air bags which fatally injure occupants, particularly children in low speed crashes, weaken the acceptability of air bags, despite their overall net safety benefits. Accordingly, to help ensure that air bags remain acceptable to the public and ultimately achieve their full potential in the future (as advanced air bags are developed and introduced), the agency believes it is reasonable and appropriate to give persons in risk groups the opportunity to obtain and use an on-off switch, upon the making of the requisite certifications on the agency request form and obtaining agency approval for each request.

The potential savings and savings foregone are described in the executive summary of the Final Regulatory Evaluation (FRE). The following discussion is based on that summary:(49)


The Final Regulatory Evaluation analyzes the potential impact of allowing motor vehicle dealers and repair businesses to install air bag on-off switches in vehicles. This option is being considered in response to concerns that current air bags may injure or kill some occupants in low speed crashes.

Data indicate that only a small portion of vehicle occupants are actually at risk of fatal harm from air bags, and that these occupants tend to fall into well-defined groups. Because both the actual risk and the public's perception of this risk are quite different for drivers and passengers, this analysis addresses each occupant position separately.

On-off switches will not be necessary after advanced air bags become available. Vehicle manufacturers are expected to install some kind of advanced air bags throughout their fleet by the year 2002. An analysis was therefore performed of the impacts that might occur during the 1998-2001 period, when an average of 45 percent of the on-road vehicle fleet will have driver air bags, and 32 percent will have passenger air bags. Safety impacts will continue to occur over the remaining life of these pre-2002 model year fleets, but at a declining rate as more vehicles are retired from the fleet without being replaced by on-off-switch-equipped vehicles. For the purposes of isolating and analyzing the impacts of this rulemaking, it is assumed that there is no change in air bag design, i.e., the potential impact of depowering or other design changes are not included. It is also assumed that there is no change in driver/passenger behavior, belt use, child restraint use, or the percent of children sitting in the front seat. Since the agency has significant education and labeling efforts underway, and the manufacturers are constantly improving air bags, the population which could be positively affected by retrofit on-off switches is actually smaller than that assumed for the purpose of this analysis. The results of this analysis are as follows:


If on-off switches are installed and used by all drivers actually at risk, the switches could prevent 45 fatalities during the 1998-2001 period, an average of 11 each year. For every one percent of those not in a risk group who always use on-off switches to turn off the driver air bag, the number of drivers saved by air bags would be reduced by 42 for that period, an average of 11 drivers each year. Nonfatal injuries impact a broad range of occupants for which particular risk groups cannot be properly identified.(50) For each one percent of drivers who always use on-off switches to turn off the driver air bag, a net increase of 490 moderate to critical injuries would occur during 1998-2001 (123 annually).(51)


Passenger impacts vary dramatically by age group. If on-off switches are always used for all child passengers (ages 0-12), they could prevent 177 deaths over the 1998-2001 period, an average of 44 deaths annually. The vast majority of these benefits would come from infants and from children 1-12 years old who ride completely unbelted, remove their shoulder belt, lean forward or otherwise place themselves at risk. The net impact of on-off switches on nonfatal injuries is uncertain, but the agency believes that on-off switches would provide a net benefit to children.

The agency cannot identify the teenage and adult at-risk group, with the exception of a minimal number of medical cases. The agency advises all those passengers above 12 years of age to leave air bags on. For every one percent of teenage and adult passengers who always utilize on-off switches to turn off their air bag, 9 additional fatalities and 93 additional moderate to critical injuries would occur, an average of 2 more fatalities and 23 more injuries annually.


NHTSA estimates that an on-off switch for one seating position would cost between $38 and $63 and that the cost for an on-off switch to control both the driver and right front passenger air bags would cost between $51 and $76 (1996 dollars) to install on aftermarket vehicles. These costs would be voluntary and incurred at the initiative of the vehicle owner. Ford was the only commenter on costs. Ford estimated the cost of installing an aftermarket on-off switch that controls both the driver and right front passenger air bag to be $95 to $124.


NHTSA notes that one commenter, MBS, submitted an analysis suggesting that a final rule would result in a large annual number of additional deaths by the year 2000. After reviewing MBS' analysis, the agency concludes that it rests on a number of incorrect assumptions about key matters and consequently cannot reliably assess the impacts of this final rule. First, MBS' analysis assumes the final rule would authorize deactivation, which is permanent and eliminates air bag protection for all vehicle users, instead of on-off switches. As noted above, on-off switches make it possible to leave air bags on except when a person at risk is riding in the vehicle. Second, MBS' analysis assumes that anyone may have their air bag turned off, based on informed decisionmaking alone. In fact, the final rule is based on informed decisionmaking, certification of risk group membership, and agency approval of each request. As a result, the final rule will reduce inappropriate requests for on-off switches, i.e., those requests based on reasons other than safety risk. Third, MBS' analysis relies on highly speculative assumptions about the percentage of respondents to telephone surveys (the January IIHS survey and a later survey by Ford) who will actually go to their dealers or repair business and purchase an on-off switch. Given the shortcomings of those early surveys, which are detailed above, they do not provide a reliable basis for estimating the level of interest in on-off switches. Although the more recent (August) survey by IIHS avoided those shortcomings and demonstrated the potential for education to reduce interest in on-off switches, that survey too does not provide a basis for reliably estimating the number of people who will obtain on-off switches under this final rule. Even though the new survey introduced key information about cost and safety, it did so only to the very limited extent that it was reasonable and practicable to do so in the context of a brief survey. Only the barest of facts were given to the respondents. Further, since IIHS was conducting a opinion survey, not a public education campaign, its efforts to educate respondents about who is at risk from air bags was very cursory. The public education campaign planned by the agency and other interested parties will provide the public with a much fuller description of the facts and present those facts in the context of persuasive explanatory discussions and graphics. Third, instead of using data representing the passenger vehicle fleet in 2000, MBS incorrectly used NHTSA data representing a later fleet fully equipped with driver and passenger air bags. By contrast, only 47 percent of the vehicles in the 2000 fleet will have driver air bags and 35 percent will have passenger air bags. The effect of this error was to magnify greatly MBS's estimate of the effects of a final rule.








49. The agency notes that IIHS and BMW raised the possibility in their comments that use of on-off switches could lead to increased occupancy of the front seat, especially by children, and thus to increased injuries and fatalities. The extent to which this phenomenon might occur, if at all, is speculative and therefore not quantifiable.

50. Some nonfatal injuries are unrelated to the factors (sitting distance from air bag and medical conditions) which define the driver risk groups. For example, since all drivers must hold the steering wheel, they are all subject to arm injuries without regard to those factors.

51. This potential increase applies to all drivers, not just those in a risk group.


[Back to section 9]   [Table of Contents]   [Next Section]