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

Seat belt use technologies currently being introduced in passenger vehicles to induce greater belt use are targeted toward the approximately 25 percent of U.S. drivers and front-seat passengers who are observed not to be buckled up (Glassbrenner 2002, 1). In this chapter, what is known about the nonuser population and its various subgroups is reviewed. The literature and recent surveys on the characteristics associated with seat belt use are summarized, and the reasons and attitudes that underlie nonbuckling behavior are examined. In the final section, inferences are drawn concerning the potential for seat belt use technologies to induce the nonuser population to buckle up.

The National Highway Traffic Safety Administration (NHTSA) conducts two surveys on seat belt use in the United States. The National Occupant Protection Use Survey (NOPUS), a probability-based observational survey of belt use by drivers and front-seat occupants of passenger vehicles, has been conducted annually since 1994. This survey provides nationally representative data on observed belt use in passenger vehicles and some demographic detail, such as belt use by race, ethnicity, and gender (Glassbrenner 2002, 13).¹ The companion Motor Vehicle Occupant Safety Survey (MVOSS), a telephone survey that provides self-reported information on belt use, has been conducted biennially since 1994.

NHTSA bases its estimates of national belt use on the observational data collected by the NOPUS. However, the survey is limited to observations of drivers and other front-seat occupants during daylight hours

and thus is not necessarily representative of high-risk driving times when belt use may be lower (Glassbrenner 2002, 18). The NOPUS can distinguish only two groups—those who are wearing their belts at the time of observation and those who are not. Thus, observed users comprise full-time users and some part-time users who are buckled up at the time of observation. Observed nonusers comprise a mix of part-time users, who for whatever reason were not buckled up at the time of observation, and habitual nonusers.

In comparison, the MVOSS can distinguish many more belt use and nonuse categories through self-reported responses to the survey questions. However, NHTSA does not consider the MVOSS a good indicator of national belt use. Self-reported belt use rates from the telephone survey are typically about 10 percentage points higher than from the NOPUS (Glassbrenner 2002, 13). The difference reflects the well-established tendency for survey participants to give socially desirable rather than completely truthful answers. It also reflects a large number of part-time respondents, who typically consider themselves to be belt users. Nevertheless, the MVOSS is the only source of unobservable demographic and socioeconomic detail about belt nonusers and insights into the reasons why motorists do not always buckle up (Glassbrenner 2002, 13).

Both the NOPUS and the MVOSS were analyzed for this study to determine what is known about the target group for seat belt use technologies—in particular, the size and characteristics of various nonuser groups. In addition, the literature on seat belt use was reviewed, and interviews and focus groups were conducted by NHTSA especially for this study to explore motorists’ behaviors and attitudes concerning belt use as well as reactions to seat belt use technologies.²

The most recent NOPUS (2002) reported a 75 percent observed national belt use rate for drivers and front-seat occupants of passenger vehicles (Glassbrenner 2002, 1). Belt use rates vary widely by state. Washington, California, Puerto Rico, and Hawaii reported observed belt use rates of

90 percent or higher, while Massachusetts recorded an observed belt use rate of 51 percent, the lowest reported (Glassbrenner 2003b, 2). Belt use rates also differ by vehicle type. The 2002 NOPUS observed belt use rates of 77 percent for passenger vehicles and 78 percent for vans and sport utility vehicles. Belt use rates for pickup trucks lagged at only 64 percent (Glassbrenner 2002, 8).

The vast majority of drivers (83 percent) interviewed in the most recent MVOSS (2000) reported wearing their seat belts “all the time” while driving. Another 9 percent reported wearing their belts “most of the time” (Block 2001, 12). However, closer investigation found inconsistencies in the responses, suggesting that the categories of belt use are more fluid than the descriptors imply. For example, 8 percent of those reporting “all-the-time” use (6.64 percent of all respondents) immediately stated in a follow-up question that they had not worn their seat belts while driving at some time during the past week (Block 2001, 24). Four percent of drivers reported wearing their seat belts “some of the time,” while few drivers acknowledged wearing their belts “rarely” (2 percent) or “never” (2 percent) (Block 2001, 12).

The literature on seat belt use typically distinguishes between belt users and nonusers. For the present study, belt users have been grouped into three broad classifications—full-time users, part-time users, and hard-core nonusers. Understanding the relative size and possible differences in attitudes and belt use behavior among different groups of nonusers in particular—the target group for seat belt use technologies—is important to determine their potential receptivity to the new technologies. On the basis of this classification system and data from the most recent MVOSS, 76 percent of drivers can be classified as full-time belt users, 20 percent as part-time users, and 4 percent as hard-core nonusers.³

Belt use by rear-seat occupants is not collected by the NOPUS. However, the MVOSS provides self-reported data on rear-seat belt use.

Whereas 83 percent of respondents reported wearing their belts “all the time” while driving and 80 percent reported wearing belts as front-seat passengers, only 49 percent reported buckling up when they sat in the rear seat (Block 2001, iv).

According to the most recent MVOSS, the vast majority (87 percent) of respondents strongly agreed that they would want to be wearing a seat belt in a crash (Block 2001, 91). Even among hard-core nonusers, more than half (56 percent) strongly or somewhat agreed that they would like to be belted in a crash (Block 2001, 92). Similar attitudes were found in the NHTSA interviews and focus groups conducted for the current study. All part-time users and 67 percent of the hard-core nonusers agreed that they greatly or somewhat reduced their risk of injury by wearing a seat belt.⁴ (Respondents to the NHTSA interviews and focus groups were classified into three groups. Full-time belt users were identified as those who responded that they forgot to wear their seat belts only once or twice or never in the past month. Hard-core nonusers reported never using a seat belt in the past month. All other respondents were classified as part-time users.) Despite positive attitudes toward belt use, many drivers and occupants continue to ride unbelted. Reasons for not using belts stem from a complex mix of habitual, situational, and attitudinal factors.

Seat belt use is often characterized as a habitual behavior rather than a conscious choice (Calisir and Lehto 2002, 802). Drivers simply follow rules they have developed on the basis of experience, rather than continuously comparing risks against benefits in deciding whether to buckle up. An individual may be triggered to buckle up by sitting in the car or driver’s seat, or by some other aspect of driving (Harrison et al. 2000, 20). Similarly, nonusers have failed to develop belt-wearing habits or have

developed a habit of nonuse (Harrison et al. 2000, 20). The habit of wearing a seat belt is learned and can be influenced by the behavior of others, including parents, peers, and children (Harrison et al. 2000, 19; Shinar 1993, 754).

Belt use may also be situational. This is particularly characteristic of part-time users, who may be cued to buckle up in some driving situations but not in others (Harrison et al. 2000, 20). Many part-time users interviewed by NHTSA for the current study and in earlier focus groups reported that they did not wear seat belts in what they considered low-risk situations (Bentley et al. 2003, 18; Bradbard et al. 1998, 12). These included short trips on familiar roads at relatively low speeds. However, these situational users tended to buckle up in poor driving conditions, such as bad weather; on longer trips involving high-speed driving on Interstates; and under congested conditions where other drivers could pose a danger (Bentley et al. 2003, 19–20).

Belt use behavior may also stem from attitudes and beliefs. Nonuse of seat belts has been related to risk-taking and other problem behaviors, such as substance abuse (Wilson 1990, 175). Many hard-core nonusers object to being forced to buckle up, believing that belt use should be a matter of personal choice (Bentley et al. 2003, 20).

Belt use is also affected by ease of use and comfort of the belt system. For example, pressure or pain from seat belts (e.g., the belt is too tight or it chokes) was reported in the most recent MVOSS as the most common complaint among those who disliked seat belts or found them annoying, particularly among women (Block 2001, 8). Fortunately, measures to improve the ease and comfort of belt use have been introduced in the passenger vehicle fleet. For example, in briefings and correspondence with the committee, two major suppliers of seat belts—Breed Technologies and Autoliv North America—noted that seat belts with height and tension adjustors and improved belt access and buckling mechanisms are already on the market.

The importance of habit, situation, attitudes, and comfort for belt use is borne out by the MVOSS and the NHTSA interviews conducted for the current study. In the most recent MVOSS (Block 2001, 62), the following were the most frequent reasons reported by drivers for not wearing a seat belt:

• Driving a short distance (59 percent),

• Forgetting to buckle up (53 percent),

• Being in a rush (41 percent), and

• Discomfort from the seat belt (33 percent).

Similar reasons for not using seat belts were reported in the NHTSA interviews. Drivers cited forgetfulness and laziness as important reasons for not buckling up. When probed, these respondents made it clear that this behavior was particularly evident when they were going on short trips or driving in familiar circumstances or at low speeds (Bentley et al. 2003, 18). Some explained that under these conditions they are not as focused on driving and tend to forget to buckle up. Others indicated that they did not see the need for buckling up in what they perceive as low-risk driving conditions (Bentley et al. 2003, 18).

A much smaller group of hard-core nonusers reported negative attitudes toward seat belts as the primary reason for nonuse in the most recent MVOSS. These reasons include discomfort, concerns that belts are dangerous in a crash (e.g., could trap the driver in the vehicle), infringement of personal freedom and resentment of authority, and the attitude that they “just don’t feel like wearing them” (Block 2001, 69–70). Similar attitudes, particularly the importance of personal choice in the decision to use a seat belt, were evident among the small group of hardcore nonusers interviewed by NHTSA for the present study (Bentley et al. 2003, 20).

The literature review conducted for this study on characteristics of seat belt use identified numerous demographic and socioeconomic characteristics associated with belt use behavior. Many of the studies are based on observational surveys and, as such, do not differentiate between different nonuser groups, such as part-time users and hard-core nonusers. Thus, as noted earlier, the user group comprises full-time and part-time users, while the nonuser group combines part-time users and nonusers.

In general, females are more likely to be observed wearing seat belts than are males. The 2002 NOPUS observed a statistically significant 7 percentage point gender difference. Females were observed using belts 79 percent of the time compared with 72 percent of the time for males (Glassbrenner 2003a, 3). A North Carolina survey of seat belt use following a high-visibility “Click It or Ticket” public information and enforcement campaign found that observed unbelted drivers were more likely to be male than observed belted drivers (Reinfurt et al. 1996, 211). Another study of primary and secondary belt use laws in four U.S. cities [Boston (secondary), Chicago (secondary), Houston (primary), and New York (primary)] observed that male drivers were less likely to buckle up than were female drivers, even in states with primary seat belt use laws (Wells et al. 2001, 5).

Generally, an individual’s age is considered to have a positive impact on belt use; older individuals are more likely to buckle up. For example, reported “all the time” belt use in the most recent MVOSS was lowest among respondents aged 21 to 24 and highest among those aged 65 and older (Block 2001, 17). An observational survey of belt use at 12 high schools in Connecticut and Massachusetts confirmed findings from earlier studies that teenagers have low belt use rates relative to other age groups, even when they drive with their parents (Williams et al. 2001).⁵

Other individual characteristics associated with age (e.g., the structure of the individual’s family), however, may confound the relation between age and belt use. One study, which compared seat belt use rates measured by observational surveys at preselected sites in Ohio with their demographic and socioeconomic characteristics as measured by U.S. Census Bureau data, did not find a strong positive correlation between age and belt use. One explanation is that older individuals may be less likely to use belts regularly when they are not living with children in

the home. Older parents with children at home may be more likely to buckle up because of their desire to set a positive example or because of pressure from children who have been exposed to public information and education media campaigns (Shinar 1993, 754). At the same time, young people (aged 18 to 24) were found to be more responsive to seat belt use when living within the traditional family structure (Shinar 1993, 754). This finding was confirmed by the most recent MVOSS. Among younger drivers who responded to the survey (aged 16 to 24), 65 percent reported that their habit of buckling up was instilled by their parents (Block 2001, vi).

Socioeconomic status is also an important factor in belt use. For example, telephone surveys conducted after a “Click It or Ticket” campaign in North Carolina found that college graduates were more likely to report driving belted than blue collar or service workers (Reinfurt et al. 1996, 213). The study of belt use laws in Boston, Chicago, Houston, and New York also confirmed through driver interviews that higher educational attainment is a strong correlate of higher seat belt use (Wells et al. 2001, 8). Education level is frequently used in the literature as an indication of socioeconomic status.

Recent studies of seat belt use among different racial groups underscore the importance of an individual’s socioeconomic status in determining the likelihood of buckling up. Racial differences alone have not proved to be important predictors of observed belt use (Reinfurt et al. 1996, 212; Glassbrenner 2003a, 13). When race is considered with other characteristics, such as gender, education, and enforcement type (i.e., primary versus secondary belt use laws), these other factors are stronger correlates of belt use (Shinar 1993, 754; Wells et al. 2001, 8). For example, the study of belt use in four cities cited above (Wells et al. 2001) observed no differences in belt use by race or ethnicity in primary law locations (p. 1). However, in secondary law locations, blacks were less likely to be belted among populations both with and without college degrees, confirming the results of other studies that show greater sensitivity to enforcement among black drivers (Wells et al. 2001, 1).

The most recent MVOSS suggests that different racial and ethnic groups may have very different perceptions about the efficacy of seat belts

that could also influence their use. The survey reported that whereas only one-third of whites agreed that seat belts were “just as likely to harm them as to help them,” almost half of blacks (46 percent) believed this statement to be true (Block 2001, 107).⁶ Forty-eight percent of Hispanics thought that seat belts were “just as likely to harm them as to help them” compared with one-third of non-Hispanics. Blacks were about twice as likely as whites to agree that putting on a seat belt made them worry more about being in a crash and were most likely to agree with the fatalistic notion that “if it was your time to die, you’ll die,” so that wearing a seat belt does not matter (Block 2001, 108).⁷ Similar findings were reported for Hispanics versus non-Hispanics. When education level is considered without regard to race or ethnicity, however, individuals with more schooling tended to be less fatalistic, less ambivalent about the injury reduction benefits of seat belt use, and less self-conscious about going against group nonbuckling norms (Block et al. 2001, vii).

Belt nonuse often is associated with a risky lifestyle, aggressive behavior, and irresponsible attitude (Wilson 1990, 176). Indeed, on the basis of self-reports, individuals who did not use seat belts or who used them inconsistently reported that they engaged in more behaviors that increase the risk for a crash, including consuming more alcohol and drugs and accumulating more traffic violations, than regular belt users (Wilson 1990, 175). A telephone interview of North Carolina motorists cited for not using seat belts found that crash rates for violators were nearly double those of a random sample of North Carolinians, and the researchers concluded that nonusers are a high-risk crash group (Williams et al. 1997, 71).

In another observational study of North Carolina drivers, unbelted drivers were significantly more likely than belted drivers to have had at

least one conviction for a traffic violation and to have been involved as a driver in at least one fatal or injury crash during the most recent 4-year period (Reinfurt et al. 1996, 212). A follow-up survey revealed that nonusers were also less likely than users to report having health coverage, more likely to acknowledge having consumed large amounts of alcohol in the past year, and more likely to have an arrest record (Reinfurt et al. 1996, 209). Other studies of observed and self-reported seat belt use have confirmed from driver records that unbelted drivers have more traffic convictions and more crashes than those who were belted (Hunter et al. 1993, 545; Preusser et al. 1991, 475).

Independent of individual demographic and socioeconomic characteristics, belt use is higher in states with primary belt use laws than in states with secondary belt use laws or in New Hampshire, which has no seat belt law (Dinh-Zarr et al. 2001, 54). The 2002 NOPUS confirms the importance of primary seat belt use laws. In primary law states, belt use rates were 80 percent. In secondary law states, belt use rates were only 69 percent, a statistically significant difference (Glassbrenner 2002, 5). Moreover, in those states that changed from a secondary to a primary belt use law, seat belt use rose (and fatalities declined) (Dinh-Zarr et al. 2001, 54). For example, when Washington State recently changed from a secondary to a primary law state, observed belt use rates rose from 83 percent in 2001 to 93 percent in 2002 (Glassbrenner 2003b, 1).

The implementation of primary enforcement laws may have a greater impact on black motorists than white motorists. In North Carolina, a primary law state, observed belt use was significantly higher among whites than blacks before implementation of the law. Since its enactment, observed belt use among blacks has exceeded use among whites (Reinfurt 2000 in Wells et al. 2001, 8). The apparent reason is that blacks perceive that they are more likely to get a ticket for belt nonuse than whites. Other studies in Louisiana, Georgia, and Maryland indicate that blacks are more sensitive to primary belt use laws because they believe that there will be a race differential in their enforcement (Solomon et al. 2000 in Wells et al. 2001, 9). Although the perception may be that blacks are targeted as offenders of primary belt use laws, studies in several states

that changed from secondary to primary laws show either no difference in the rate of ticketing between blacks and whites or a relative increase in the ticketing of whites after the enactment of a primary law (Dinh-Zarr et al. 2001, 54).

Differential enforcement is undesirable. However, the perception that laws are being strictly enforced makes them work (Wells et al. 2001, 9). According to the most recent MVOSS, the percentage of Americans who thought that ticketing for seat belt nonuse was an imminent threat was significantly higher in primary law states than in secondary law states (Block 2001, ix).

A review of the literature, survey data, and interview results suggests that the reasons for seat belt use and nonuse are complex. Age, gender, vehicle type, and enforcement level (i.e., primary versus secondary law states) have all been shown to affect belt use. Lower belt use is associated with young adults, males, pickup trucks, and states with secondary belt use laws or no law, like New Hampshire. These simple correlates of belt use, however, are confounded by other variables, such as education level, and by situational and attitudinal factors.

For purposes of this study, differences among nonuser groups are important for assessing the likely impact of seat belt use technologies. The literature is sparse, however, concerning the factors differentiating nonusers—the target group for seat belt use technologies. The available survey data, primarily from the MVOSS, suggest that there are at least two nonuser groups: part-time users (those who buckle up less than all the time) and hard-core nonusers (those who never buckle up).

Part-time users appear to be the predominant nonuser group. Members of this group generally express positive attitudes toward seat belts but do not always buckle up. Many appear not to have developed the habit of wearing a belt and thus forget to buckle up. Others choose to use belts only in situations of perceived risk—long trips at high speeds on unfamiliar roads. Part-time users should be amenable to seat belt use technologies that help remind them to buckle up. Moreover, if reminder

systems can help part-time users develop habits of belt use, they may have a lasting impact on this segment of the nonuser population.

Hard-core nonusers are a much smaller segment of the nonuser population. However, the importance of this group should not be understated, because of its overrepresentation in fatal crashes and other high-risk driving behaviors, such as speeding and driving while impaired by alcohol. Seat belt reminder systems are likely to have little effect on hard-core nonusers who choose not to buckle up. They generally do not acknowledge the benefits of seat belts and are opposed to their use. More aggressive solutions, such as interlock systems, may be needed to get this small, but important, nonuser group to buckle up. However, will the hard-core nonusers object to such intrusive technologies? In the next chapter, the U.S. experience with interlock systems, among other technologies, is reviewed.

Bentley, J. J., R. Kurrus, and N. Beuse. 2003. Qualitative Research Regarding Attitudes Towards Four Technologies Aimed at Increasing Safety Belt Use. Report 2003-01. Equals Three Communications, Inc., Bethesda, Md., June. [National Highway Traffic Safety Administration Docket No. 13226 in the U.S. Department of Transportation’s electronic docket system (dms.dot.gov).]

Block, A. W. 2001. 2000 Motor Vehicle Occupant Safety Survey. Vol. 2, Seat Belt Report. National Highway Traffic Safety Administration, U.S. Department of Transportation, Nov.

Bradbard, S. L., J. C. Panlener, and E. Lisboa-Farrow. 1998. Increasing Seat Belt Use Among Part-Time Users: Messages and Strategies. DOT-HS-808-708. National Highway Traffic Safety Administration, U.S. Department of Transportation, Feb.

Calisir, F., and M. R. Lehto. 2002. Young Drivers’ Decision Making and Safety Belt Use. Accident Analysis and Prevention, Vol. 34, No. 6, Nov., pp. 793–805.

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.

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.

Glassbrenner, D. 2003a. Safety Belt Use in 2002—Demographic Characteristics. Research Note. DOT-HS-809-557. National Highway Traffic Safety Administration, U.S. Department of Transportation, March.

Glassbrenner, D. 2003b. Safety Belt Use in 2002—Use Rates in States and Territories. Research Note. DOT-HS-809-587. National Highway Traffic Safety Administration, U.S. Department of Transportation, May.

Harrison, W. A., T. M. Senserrick, and C. Tingvall. 2000. Development and Trial of a Method to Investigate the Acceptability of Seat Belt Reminder Systems. Report 170. Monash University Accident Research Centre, July.

Hunter, W. W., R. J. Stewart, J. C. Stutts, and E. A. Rogman. 1993. Observed and Self-Reported Seat Belt Wearing as Related to Prior Traffic Accidents and Convictions. Accident Analysis and Prevention, Vol. 25, No. 5, pp. 545–554.

Preusser, D. F., A. F. Williams, and A. K. Lund. 1991. Characteristics of Belted and Unbelted Drivers. Accident Analysis and Prevention, Vol. 23, No. 6, pp. 475–482.

Reinfurt, D. W. 2000. Memo to the Director of the Governor’s Highway Safety Program. University of North Carolina Highway Safety Research Center, Chapel Hill, Nov.

Reinfurt, D., A. Williams, J. Wells, and E. Rogman. 1996. Characteristics of Drivers Not Using Seat Belts in a High Belt Use State. Journal of Safety Research, Vol. 27, No. 4, pp. 209–215.

Shinar, D. 1993. Demographic and Socioeconomic Correlates of Safety Belt Use. Accident Analysis and Prevention, Vol. 25, No. 6, pp. 745–755.

Solomon, M. G., L. A. Cosgrove, and D. F. Preusser. 2000. A Summary of Results from Studies Measuring the Change from Secondary Enforcement of Safety Belt Laws to Primary Enforcement Emphasizing the Effects on Race. Presentation at the National Highway Traffic Safety Administration Restraint Use Strategy Workshop, Crystal City, Va.

Wells, J., A. F. Williams, and C. M. Farmer. 2001. Seat Belt Use Among African Americans, Hispanics, and Whites. Insurance Institute for Highway Safety, Arlington, Va.

Williams, A. F., J. K. Wells, and D. Reinfurt. 1997. Characteristics and Opinions of North Carolina Residents Cited for Not Using Seat Belts. Journal of Traffic Medicine, Vol. 25, Nos. 3–4, pp. 71–76.

Williams, A. F., A. T. McCartt, and L. Geary. 2001. Seat Belt Use by High School Students. Injury Prevention, Vol. 9, pp. 25–28.

Wilson, R. J. 1990. The Relationship of Seat Belt Non-Use to Personality, Lifestyle, and Driving Record. Health Education Research, Theory and Practice, Vol. 5, No. 2, pp. 175–185.