Buckling Up: Technologies to Increase Seat Belt Use -- Special Report 278 (2004)

In this chapter, a brief review is provided of the 1970s experience with early seat belt reminder systems, the ignition interlock, and the 4- to 8-second belt reminder that remains in vehicles today. Although results from studies of the effectiveness and acceptability of these systems may not still be valid, the way in which these technologies were introduced has a continuing effect on their acceptability and likely effectiveness today. The chapter includes a review of experience with other key approaches for increasing seat belt use, primarily the enactment of seat belt use laws. The chapter ends with a brief summary of lessons learned from the past that can influence the successful introduction of seat belt use technologies today.

Early seat belt use technologies were introduced as alternatives or interim measures to the primary and preferred approach of the National Highway Traffic Safety Administration (NHTSA) for increasing seat belt use at that time—introduction of passive restraint systems.¹ On January 1, 1972, NHTSA required passenger vehicles for sale in the United States to be equipped with passive restraints protecting vehicle occupants in frontal barrier crashes up to and including 30 mph, or alternatively, with a buzzer–light reminder system. With few exceptions of cars sold with inflatable front cushions, the automobile manufacturers opted for the reminder system (Robertson 1975, 1320). The system consisted of a flashing light and buzzer, which activated continuously for at least 1 minute if the vehicle was placed in gear and the driver or front outboard passenger was not belted (Robertson 1975, 1320). The simple sensor system used to activate the reminder system, however, could be bypassed

easily. Moreover, once the belt was left in an extended position or buckled, the reminder system would not be activated again (Westefeld and Phillips 1976b, vii).

When it became evident that the introduction of passive restraint systems would be delayed, NHTSA moved to require ignition interlock systems on all cars as an interim measure. Effective August 15, 1973, NHTSA required that all Model Year (MY) 1974 passenger vehicles be equipped with an ignition interlock that allowed the vehicle to start only if the driver was seated and the belts were extended more than four inches from their normally stowed position or the belts were latched (Robertson 1975, 1320). In addition, an audible warning was activated if seat belts were unfastened during the trip.² It was hypothesized that the ignition interlock would increase seat belt use by eliminating two of the most popular ways of defeating the early belt reminder systems: leaving the belt fastened and tucking it behind the seat, or tying a knot in the belt so that it was held out of the retractor (Cohen and Brown 1973, 5).

When Congress passed legislation prohibiting NHTSA from requiring either the ignition interlock or continuous buzzer systems as described in Chapter 1, NHTSA changed Federal Motor Vehicle Safety Standard (FMVSS) 208 to a less aggressive requirement. Passenger vehicles manufactured after February 1975 were required to have a warning light of 4- to 8-seconds duration that is activated when the ignition is turned on regardless of whether the seat belt is fastened, and a chime of similar duration that sounds unless the driver’s belt is buckled (Westefeld and Phillips 1976b, viii). This standard, which so far as can be determined was specified without any empirical justification, is still in effect today.

NHTSA introduced the ignition interlock system without extensive study or pilot testing. Only one NHTSA-funded study could be found (Cohen

and Brown 1973) that had compared belt use in rental car fleets equipped with various combinations of buzzer–light and interlock systems prior to the interlock regulation.³ That study found a significant increase in belt use for equipped vehicles relative to vehicles without any reminder or interlock system. However, it showed no significant difference in belt use by drivers of cars with buzzers versus those with interlock systems (Cohen and Brown 1973, 3; Robertson 1975, 1324). Rental car drivers without any reminder or interlock system used seat belts on 23 percent of their trips compared with 51 percent and 49 percent of rental car drivers with the two reminder-only systems, and 56 percent of rental car drivers with the reminder–interlock system (Cohen and Brown 1973, 3). Substantial resistance to each of the reminder and interlock systems was observed, however, with one-third of drivers in each group claiming that they would disconnect the system if it were installed in their personal vehicles. Drivers with the more demanding systems had less favorable attitudes and were more likely to say that they would either modify or disconnect the systems (Cohen and Brown 1973, 26).

The Insurance Institute for Highway Safety had conducted an earlier study that compared belt use in MY 1972 cars equipped with the NHTSA-required buzzer–light system with belt use in unequipped cars (Robertson and Haddon 1972 in States 1973). That study, which was focused on personal vehicles, did not detect any significant effect on seat belt use for vehicles equipped with the buzzer–light system. The result, which contradicted the NHTSA study of rental car fleets (Cohen and Brown 1973) that had observed a significant difference in belt use between buzzer–light and nonequipped vehicles, can be attributed in part to the fact that rental car drivers are less likely to attempt to disable a system on the rental car than on their personal vehicle—a finding noted in the NHTSA-funded rental car study (States 1973, 435; Cohen and Brown 1973, 26).

Following the interdiction of the ignition interlock and continuous buzzer–light systems, a 1975 study (Robertson 1975) documented the

extent of belt use observed in 1972, 1973, and 1974 passenger vehicles with different seat belt use technologies. Belt use was measured at 138 sites in Baltimore, Maryland; Houston, Texas; Los Angeles, California; the New Jersey suburbs of New York City; Richmond, Virginia; and Washington, D.C. Drivers in 48 percent of MY 1974 passenger vehicles equipped with ignition interlock systems were using lap and shoulder belts, and 11 percent were using lap belts only, for a total use rate of 59 percent. Only 7 percent of drivers in MY 1973 passenger vehicles equipped with the buzzer–light systems were using lap and shoulder belts, and 21 percent were using lap belts only, for a total use rate of 28 percent. Twenty-five percent of drivers of MY 1972 passenger vehicles equipped with buzzer–light systems were buckled up, and 23 percent of drivers of MY 1972 vehicles without any reminder systems were using one or both belts. Thus, the study showed an unambiguous positive effect on belt use for the 1974 interlock-equipped passenger vehicles relative to vehicles with reminder systems or without any system.

In 1976 NHTSA reported effectiveness rates for both buzzer–light and sequential logic seat belt interlock technologies by measuring belt use rates in 19 U.S. cities (Westefeld and Phillips 1976b). The study showed that ignition interlocks were initially effective and more than doubled belt use rates from about 28 percent in MY 1973 vehicles equipped only with buzzer–light reminders to about 67 percent in MY 1974 cars equipped with ignition interlocks (Westefeld and Phillips 1976b, vii). However, initial increases in use rates from seat belt interlocks decreased over time as many motorists eventually disconnected the seat belt interlock or circumvented it, thus never developing the intended positive belt use habits.

A 1976 NHTSA study (Westefeld and Phillips 1976a) examined the effectiveness of the 4- to 8-second chime and light reminder systems on MY 1975 and 1976 vehicles that replaced the ignition interlock as well as the effectiveness of various other types of warning systems allowable in the postinterlock environment. For that study, 818 rental cars were modified with reminders that met revised FMVSS 208, which included six different reminder systems. The effectiveness data reported were based on 5,429 observations of rental car drivers at a single location: the Sky Harbor Airport in Phoenix, Arizona (Westefeld and Phillips 1976a,

1). The control group for the study consisted of drivers of vehicles with no reminder systems. Of these drivers, 12.8 percent used lap and shoulder belts leaving the rental car terminal. Of the drivers of vehicles equipped with the FMVSS 208 8-second chime and light warning systems, 13.1 percent were buckled—an insignificant difference. Thus, the authors concluded that the 8-second reminder system required by FMVSS 208 (still in place today) was not effective in increasing seat belt use (Westefeld and Phillips 1976a, 2). Moreover, the study showed that the most effective system would include, in addition to the 8-second reminder technology required by FMVSS 208, a reminder light that would not turn off until the driver buckled up, as well as a sequential logic system that required the driver first to sit down on the seat and second to buckle the belt so that the system could not be easily circumvented (Westefeld and Phillips 1976a, 2).

Since seat belts were first installed in passenger vehicles, the federal government, states, safety groups, and the private sector have tried a variety of approaches to increase seat belt use. Early efforts (late 1960s, early 1970s) to educate the public on the benefits of seat belts that relied solely on public service advertisements and major media campaigns proved largely ineffective in increasing belt use (O’Neill 2001; Haseltine 2001). NHTSA then turned to technological approaches—warning systems, ignition interlocks, and passive restraints—during the 1970s and 1980s to increase belt use. Enactment of state belt use laws, spurred by the 1984 regulation on passive restraints (details to follow) and accompanied by a massive industry lobbying campaign, resulted in significant gains in belt use by the early 1990s (Haseltine 2001). Efforts to strengthen belt use laws combined with well-publicized intensive enforcement campaigns further increased belt use rates during the 1990s (Haseltine 2001). The discussion that follows highlights a few of the key strategies.

After the interlock requirement interdiction, NHTSA turned its focus once again to passive restraint systems. A nationwide survey conducted

in 1978 suggested that the American public believed, by a two-to-one margin, that the government should require automatic crash protection in new cars rather than adopt policies that “force” behavior either by technology like the ignition interlock or by federal or state laws requiring the use of seat belts with fines for nonuse (Hart Research 1978, 4–5). At the time the survey was conducted, only one-quarter of the respondents reported that they used their seat belts all or most of the time. Among infrequent users, seat belts were seen as confining, bothersome, and uncomfortable. Many nonusers believed that seat belts posed serious safety problems (Hart Research 1978, 2–3).

The previously described Dole decision in 1984 provided for the phase-in of automatic protection in cars beginning with MY 1987 passenger vehicles; such protection would become mandatory in MY 1990 (Kratzke 1995, 7). Once the passive restraint requirement was mandated by NHTSA, the initial reaction of the industry was to develop passive seat belt systems that automatically restrained the occupant once in the vehicle. There were various types of two-point (passive shoulder belt with knee bolsters and active lap belts) and three-point motorized and non-motorized systems.

One of the earliest passive belt systems was the two-point passive shoulder belt. It was developed and introduced in 1975 in the Volkswagen Rabbit and then in mid-1978 in the General Motors (GM) Chevette. It used an automatic diagonal shoulder belt with a manual lap belt and an energy-absorbing knee bolster for those who only used the passive belt. The system had initial acceptance, but use of the manual lap belt was low due to inconvenience of access and forgetfulness of the occupant.

The most popular passive belt designs involved a motorized shoulder belt that acted on a track running along the roof rail (Johannessen 1987, 3). In the stowed position, the upper anchor of the shoulder belt was forward in the vehicle close to the steering wheel. In the restrained position, the track moved rearward on the roof rail, bringing the upper anchor of the shoulder belt rearward in position to restrain the occupant in a crash. These systems also used a manual lap belt. This approach facilitated ingress and egress while serving as a passive shoulder belt restraint. However, failure to buckle the manual belt could result in an occupant sliding out of position in a crash (e.g., “submarining” or sliding under the belt) and thus being vulnerable to injury.

GM introduced a three-point nonmotorized passive belt system in 1980 to comply with the passive restraint requirement. It was designed so that, when the door was open, the shoulder and lap belt, which were attached to the rear edge of the door, swung outward with the door. This allowed ingress when the belts were pushed forward to give access to the seat. When the door was closed, the belts moved closer to the occupant and formed the passive three-point belt system. However, the system was almost always used as an active lap–shoulder belt configuration by unlatching the belt to exit the vehicle. Despite this common practice, field studies of belt use still showed an increase in wearing rates with this door-mounted system.

Most surveys after introduction of passive belt systems found an increase in wearing rates (Johannessen 1987, 3), but the systems proved cumbersome to wear, were fraught with reliability issues, and had a required easy disconnect feature. In addition, early studies showed potential injuries with use of the automatic two-point systems when the manual lap belt was not buckled (Evans 1990). It became clear that it was difficult for the passive belt systems to match the safety performance of active lap–shoulder belts, and the industry turned to the other option in the passive requirements—air bags.

Within a few years, most vehicles were being produced with driver air bags, and the conventional lap–shoulder belt system reemerged in vehicles. The introduction of air bags is now complete, with all light vehicles equipped with driver and passenger air bags and active three-point lap–shoulder belt systems. While voluntary belt wearing has now increased to 75 percent nationally with various measures discussed elsewhere in this report, the historic issue of technological solutions to encourage belt wearing has resurfaced.

The 1984 Dole decision also encouraged states to pass mandatory seat belt use laws, with the proviso that the automatic protection requirements might be eliminated if, by April 1989, the Secretary of Transportation found that two-thirds of the nation’s population was covered by state-mandated seat belt use laws (Kratzke 1995, 8). The required threshold was not reached, but the widespread introduction of seat belt

use laws resulted in what has proved to be one of the most effective approaches for increasing belt use (Dinh-Zarr et al. 2001, 48). The automobile industry commenced a massive lobbying campaign and formed a new organization, Traffic Safety Now, to convince states to enact seat belt use laws (Haseltine 2001). By the time Traffic Safety Now closed its doors in 1992, 93 percent of the U.S. population was subject to state seat belt use requirements (Haseltine 2001).

Unfortunately, not all belt use laws are equal. Primary belt use laws allow a police officer to stop a motorist solely for not wearing a seat belt; secondary belt use laws allow a police officer to issue a seat belt citation to an unbuckled motorist only after the motorist has been stopped for another traffic violation (Glassbrenner 2002, 5). In the United States, primary belt use laws have been the exception rather than the rule. In contrast, secondary belt use laws are unknown outside the United States (Dinh-Zarr et al. 2001, 52).

According to NHTSA’s most recent National Occupant Protection Use Survey (Glassbrenner 2002, 1), the difference in belt use rates between primary and secondary law states is a statistically significant 11 percent. Belt use rates are 80 percent in the 17 states plus the District of Columbia and Puerto Rico with primary laws, but only 69 percent in the 32 states with secondary belt use laws and in New Hamphsire, the only state without any belt use law for adults (Glassbrenner 2002, 5).

Strong enforcement is a necessary component of effective seat belt use laws. Motorists must be convinced that violators will be ticketed and nontrivial penalties exacted. State-conducted studies of belt use rates and state restraint law penalties in 1998 and 1999 found that belt use rates averaged 6 percentage points higher in states with fines and court costs of $30 and above than in states with fines and court costs less than $30 (Haseltine 2001). A study of the effectiveness of a seat belt use law in Texas, a primary enforcement state, found a statistically significant reduction in driver-involved injury rates when fines were introduced for belt use violations (Loeb 1995, 84). Another study evaluating a belt use law in North Carolina, also a primary enforcement state, found that traffic injuries were reduced more when traffic citations for belt nonuse were given rather than warnings (Reinfurt et al. 1990). The 1998–1999 state survey results suggest that penalty levels can also make a difference in secondary law states (Haseltine 2001).

Stronger state belt use laws (changing from secondary to primary laws), combined with well-publicized intensive enforcement campaigns, have largely accounted for gains in seat belt use during the 1990s (Haseltine 2001). NHTSA initiated the first wide-scale effort to mobilize state and local law enforcement agencies to enforce seat belt use laws, named “Operation Buckle Down,” from 1990 to 1992 (Haseltine 2001). The comprehensive Canadian Selective Traffic Enforcement Program model, which combined training of law enforcement officials, high-visibility enforcement campaigns, and pre– and post–belt use surveys and public education efforts, was introduced in North Carolina in 1993. The 5-year statewide “Click It or Ticket” program became a model for high-visibility enforcement programs in other states (Haseltine 2001). Today, most states conduct month-long, federally supported “Click It or Ticket” seat belt campaigns, typically in May and November each year (AASHTO Journal 2003, 16).

The early experience with technologies encouraging seat belt use suggests that the ignition interlock was effective initially in bringing about large increases in seat belt use at a time when belt use rates were low generally, laws mandating belt use were unknown, belt designs had not matured, and the general public was not convinced of the safety benefits of buckling up. However, the interlock technology was intrusive, and the general public was largely unprepared for its rapid introduction. Systems were implemented without adequate field trials and with evaluations that were at best unsophisticated and at worst unreliable, which caused the public to become disenchanted with the technology. For example, by definition the ignition interlock did not allow the driver to idle or drive at low speeds without wearing a belt. Although nearly 30 years has passed since the interlock experience, both NHTSA and the automobile manufacturers remain wary of technologies that the public may find excessively intrusive. Successful technology introduction today will require more careful balancing of effectiveness and intrusiveness and more attention to studying and evaluating different technologies.

The limited study of the 4- to 8-second light and chime reminder systems that replaced the interlock and remain in effect today showed no

statistically significant effect on belt use relative to passenger vehicles without the reminder systems. Moreover, there appears to be no scientific basis for the selection of the 4- to 8-second duration of the reminder.

Much has changed since the brief experiment with interlocks nearly 30 years ago. Observed seat belt use has increased from the teens in the 1970s to 75 percent in 2002, largely because of enactment and enforcement of state seat belt use laws. While highly publicized enforcement efforts have contributed to recent increases in observed use, weak state seat belt use laws with secondary enforcement provisions and low fines hinder progress. Seat belts in modern vehicles are easier to use and more comfortable than those of the 1970s. Electronics and sensor systems in vehicles are also vastly improved. Because of these circumstances, it is appropriate and potentially fruitful to explore new vehicle technologies to assist in increasing seat belt use.

AASHTO American Association of State Highway and Transportation Officials

AASHTO Journal. 2003. U.S., DOT, NHTSA Launch “Click It or Ticket” Seat-Belt Campaign. Vol. 103, No. 20, May 16, p. 16.

Cohen, J. B., and A. S. Brown. 1973. Effectiveness of Safety Belt Warning and Interlock Systems. DOT-HS-800-859. National Highway Traffic Safety Administration, U.S. Department of Transportation, April.

Dinh-Zarr, T. B., D. A. Sleet, R. A. Shults, S. Zaza, R. W. Elder, J. L. Nichols, R. S. Thompson, and D. M. Sosin. 2001. Reviews of Evidence Regarding Interventions to Increase the Use of Safety Belts. American Journal of Preventive Medicine, Vol. 21, No. 4S, Nov., pp. 48–65.

Evans, L. 1990. Motorized Two-Point Safety Belt Effectiveness in Preventing Fatalities. Annual Meeting, Association for the Advancement of Automotive Medicine, Scottsdale, Ariz., Oct. 1.

Glassbrenner, D. 2002. Safety Belt and Helmet Use in 2002—Overall Results. DOT-HS-809-500. National Highway Traffic Safety Administration, U.S. Department of Transportation, Sept.

Hart Research. 1978. Public Attitudes Toward Passive Restraint Systems: Summary Report. National Highway Traffic Safety Administration, U.S. Department of Transportation, Aug.

Haseltine, P. W. 2001. Seat Belt Use in Motor Vehicles: The U.S. Experience. In 2001 Seat Belt Summit, Automotive Coalition for Traffic Safety, Inc., Jan. 11–13. www.actsinc.org/Acrobat/SeatbeltSummit2000.pdf.

Johannessen, G. H. 1987. Historical Review of Automatic Seat Belt Restraint Systems. SAE Technical Paper 870221. International Congress and Exposition, Society of Automotive Engineers, Detroit, Mich., Feb.

Kratzke, S. R. 1995. Regulatory History of Automatic Crash Protection in FMVSS 208. SAE Technical Paper 950865. International Congress and Exposition, Society of Automotive Engineers, Detroit, Mich., Feb. 27–March 2.

Loeb, P. D. 1995. The Effectiveness of Seat-Belt Legislation in Reducing Injury Rates in Texas. American Economic Review, Vol. 85, No. 2, May, pp. 81–84.

O’Neill, B. 2001. Seat Belt Use: Where We’ve Been, Where We Are, and What’s Next. In 2001 Seat Belt Summit, Automotive Coalition for Traffic Safety, Inc., Jan. 11–13. www.actsinc.org/Acrobat/SeatbeltSummit2000.pdf.

Reinfurt, D. W., B. J. Campbell, J. R. Stewart, and J. C. Stutts. 1990. Evaluating the North Carolina Safety Belt Wearing Law. Accident Analysis and Prevention, Vol. 22, pp. 197–210.

Robertson, L. S., and W. Haddon, Jr. 1972. Belt Warning Devices. Insurance Institute for Highway Safety, Washington, D.C.

Robertson, L. S. 1975. Safety Belt Use in Automobiles with Starter-Interlock and Buzzer-Reminder Systems. American Journal of Public Health, Vol. 65, No. 12, pp. 1319–1325.

States, J. D. 1973. Restraint System Usage—Education, Electronic Inducement Systems, or Mandatory Usage Legislation? SAE Paper 1973-12-0027, pp. 432–442.

Westefeld, A., and B. M. Phillips. 1976a. Effectiveness of Various Safety Belt Warning Systems. DOT-HS-801-953. National Highway Traffic Safety Administration, U.S. Department of Transportation, July.

Westefeld, A., and B. M. Phillips. 1976b. Safety Belt Interlock System Usage Survey. DOT-HS-801-957. National Highway Traffic Safety Administration, U.S. Department of Transportation, Aug.