Seat belt use in top-grossing movies vs actual US rates, 1978-1998.

Abstract

The prevalence of risky health behaviors, such as tobacco and alcohol use, is consistently higher in movies than in real life.1–6 We compared the prevalence of a protective health behavior—seat belt use—in movies and in reality. Two hundred eleven movies (approximately 10 per year) were randomly selected from top-grossing releases (N = 917) from 1978 through 1998. Movies that took place outside the United States or in noncontemporary time relative to the release date, movies that contained no humans, movies that had X or NC-17 ratings (on September 27, 1990, the Motion Picture Association of America changed the name of the X category to NC-17: no one 17 and under admitted), and cartoons were excluded from the sampling frame. Each movie was coded for seat belt use, occupant gender, movie rating, and movie genre. Front seat occupants were classified as consistent users if they wore a seat belt every time they appeared in a moving vehicle (interrater agreement = 93%). Inconsistent use and nonuse also were recorded, with inconsistent users evenly assigned at random to either consistent or nonuser status for analyses. The rate of seat belt use in movies did not exceed 10% until 1987, peaked that year at 32%, and has fluctuated between 10% and 30% since then. National Highway Traffic Safety Administration (NHTSA) data from the same period indicate low rates of seat belt use in the United States before 1984, but the rate increased 3-fold in the mid-1980s following passage of seat belt laws in 31 states.7 As enforcement and public education were strengthened, rates increased further (see Figure 1 ▶).7 FIGURE 1— Actual and movie rates of seat belt use, by year: United States,1978–1998. In general, increases in NHTSA rates preceded similar, but smaller, increases in movie rates. A cross-correlation time-series analysis suggested that NHTSA rates are a leading, not lagging, indicator of movie rates. NHTSA rates for a given year were most highly correlated with movie rates from 1 year later (r = 0.404, P < .10). Seat belt use was higher in family movies than in other movies (23% vs 11%, P = .01), but otherwise neither movie genre nor rating was significantly associated with seat belt use. In movies that showed both men and women as front seat occupants (n = 49), women were more likely than men to wear a seat belt (18% vs 10%; t test = –2.676, P < .01). Clear differences exist between actual and movie rates of seat belt use. Less clear is what influence, if any, movies have on viewers' attitudes and behaviors. Others have suggested that risky health behaviors shown in movies influence audience members directly, through modeling.1–6 In the case of protective health behaviors, we are skeptical of a direct effects explanation. Rather, we think that the cumulative effect of seeing little or no seat belt use in movies may be to distort viewers' perception of social norms. Although some driving scenes in movies are filmed in unmoving vehicles on studio sets, we know of no reason that this or other conventions of filmmaking would require nonuse of seat belts. Nor do we know of any interests served by intentionally minimizing seat belt use in movies. This lack of seat belt use calls for a response from the motion picture industry. Every time a character is shown in a moving vehicle without wearing a seat belt, an opportunity—however subtle—is missed to depict seat belt use.