often than women in accidents caused by speeding and driving under the in¯uence of alcohol,women were more frequently involved in accidents caused by judgment errors. Harre, Field, andKirkwood (1996) found that men, more than women, engage in unsafe driving behaviors, such asdriving after drinking and speeding.Age is another demographic variable frequently found to be related to risky driving. Youngerdrivers violate the law more often than do older drivers (Groeger & Brown, 1989; Jonah &Dawson, 1987; Parker, Reason, Manstead, & Stradling, 1995), are more involved in crashes(Evans, Wasielewski, & Von-Buseck, 1982) and suer more fatal road accidents (Arnett, 1990;Levy, 1990).Several studies have found an interactive eect of gender and age on driving behavior: Youngmale drivers are considered a high-risk group in regard to accident involvement (Arnett, 1990),risky driving (Groeger & Brown, 1989), aggressive driving (Simon & Corbett, 1996), violation of trac laws (Jonah & Dawson, 1987) and even parking illegally in spaces reserved for people withdisabilities (Fletcher, 1995).
1.1. Attitudes toward trac laws
Perception of the danger involved in the commission of trac violations has frequently beendescribed as aecting driving behavior (Dejoy, 1992; Finn & Bragg, 1986; Matthews & Moran,1986; Tra Ènkle, Gelau, & Metker, 1990). Driving behavior, however, is likely to be in¯uenced by amore comprehensive system of drivers' attitudes which is described in models of social in¯uence.Tyler (1990) presents a dierentiation in regard to instrumental and normative motives forcompliance with the law. Instrumental motives are related to the gains and losses involved inobeying or disobeying the law. In the area of driving, losses are the danger of a road accidentresulting from the commission of violations or the risk of apprehension (Shinar & McKnight,1986). The gains involved in driving are pleasure and convenience (Arnett, 1990; Rothengatter,1988; Rutter, Quine, & Chesham 1995). Normative motives result from the internalization of thelaw and the perceived legitimacy of the authorities enforcing the law (Tyler, 1990). Kelman (1961)describes reactions to in¯uence and dierentiates between compliance which is initiated by adesire to avoid punishment or to receive positive rewards, and the internalization of an attitudebecause it is perceived as coherent with reality as well as the individual's general system of valuesand beliefs. Compliance is achieved through control, and is expressed only in the presence of thein¯uence agent. On the other hand, internalization results in a long lasting eect of the attitudewhich does not depend on the presence of the in¯uence agent.
1.2. Rational choice theory explanations for gender and age in¯uence
The rational-choice theory of oending explains crimes in terms of the costs and bene®ts of committing violations (Cornish & Clarke, 1986) and predicts that the intention to commit illegalbehavior is inversely related to the perceived costs of the act. In the area of trac violations, thisexplanation is suggested for gender dierences in driving behavior. Studies show that male driversunderestimate the hazards involved in various driving activities (Dejoy, 1992) and assess theirdriving ability more highly than do female drivers assess theirs (Dejoy, 1992; Matthews & Moran,1986). For example, McKenna, Stanier, and Lewis (1991) found that men tended to rate their
D. Yagil/Transportation Research Part F 1 (1998) 123±135